Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides

UOPP #1290994, VOL 49, ISS 2

Patrizia Ferraboschi, Samuele Ciceri, and Paride Grisenti

QUERY SHEET This page lists questions we have about your paper. The numbers displayed at left can be found in the text of the paper for reference. In addition, please review your paper as a whole for correctness.

There are no Editor Queries in this paper.

TABLE OF CONTENTS LISTING The table of contents for the journal will list your paper exactly as it appears below: Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides Patrizia Ferraboschi, Samuele Ciceri, and Paride Grisenti Organic Preparations and Procedures International, 49:1–85, 2017 Copyright Ó Taylor & Francis Group, LLC ISSN: 0030-4948 print / 1945-5453 online DOI: 10.1080/00304948.2017.1290994

Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides

Patrizia Ferraboschi, Samuele Ciceri, and Paride Grisenti Dipartimento di Biotecnologie Mediche e Medicina Traslazionale, Universita 5 degli Studi di Milano, Via Saldini 50, 20141 Milano, Italy

Introduction Nucleosides, due to their biological role as constituents of nucleic acids, are main targets in the development of analogues aimed at antimetabolite-based therapy. Modified nucleosides can disrupt biological processes causing the death of cancer or virally-infected cells. 10 Fluorinated analogues of biologically active compounds are often characterized by a dramatic change in their activity, compared with the parent molecules. Fluorine, the most electronegative element, is isosteric with a hydroxy group, the C-F bond length (1.35 A) being similar to the C-O bond length (1.43 A). In addition, it is the second smallest atom and it can mimic hydrogen in a modified structure; its van der Waals radius (1.47 A) is intermedi- 15 ate between that of hydrogen (1.20 A) and that of oxygen (1.52 A). The strength of the C-F bond exceeds that of C-H bond and for this reason organofluorine compounds are often biologically and chemically more stable than their corresponding natural compounds. In the case of nucleosides and their analogues, fluorine atoms can be introduced either in the nucleobase or in the sugar moiety. An example of the first type modification 20 is capecitabine1 (N4-pentyloxycarbonyl-50-deoxy-5-fluorocytidine), a 5-fluoropyrimidine nucleoside approved as a drug against colorectal, gastric and breast tumors; gemcitabine2 (20-deoxy-20,20-difluorocytidine) is an example of a nucleoside fluoro-modified in the sugar moiety, approved as a drug against solid tumors. The aim of this review is to discuss the synthesis of antitumor pyrimidine nucleosides 25 containing fluorine atoms in either the nucleobase or the sugar moiety. Because of the ongoing need of new antitumor chemotherapies, over the years many fluorinated pyrimidine nucleosides have been prepared in order to assay their activity. The synthesis of these compounds has thus been driven largely by results on their biological potential. Nonetheless, it is important for experimentalists to be aware of the full range of methods 30 used, whether or not the compounds synthesized have actually demonstrated antitumor activity. It is our hope, then, that researchers investigating fluoropyrimidines for purposes other than their anticancer properties will also find this article useful.

Received June 1, 2016; in ﬁnal form November 15, 2016. Address correspondence to Patrizia Ferraboschi, Dipartimento di Biotecnologie Mediche e Medicina Traslazionale, Universita degli Studi di Milano, Via Saldini 50, 20141 Milano, Italy. E-mail:[email protected] This work is dedicated to Dr Giuseppe Celasco deceased on May 10, 2016.

1 2 Ferraboschi, Ciceri, and Grisenti

I. Fluoropyrimidines

1. 5-Fluorouracil 35 Interest in the fluoropyrimidines stemmed from studies of the metabolism of uracil in rat hepatoma cells. The observation that these cells utilize uracil more avidly than normal rat intestinal mucosa prompted the preparation of fluorinated pyrimidines in order to improve disruption of tumor DNA biosynthesis.3 In 1957 5-fluorouracil (5-FU) 1 was synthesized by Heidelberger et al.,4 with the aim of 40 blocking metabolism in malignant cells. The replacement of a hydrogen atom at C-5 by the fluorine atom modifies the interaction with the active sites of enzymes involved in metabolism. This antimetabolite, although toxic, is still one of the most widely used agents against solid tumors. Its action is due to two different mechanisms:5 after penetration into the cell, 5-FU 1 is transformed into the 5-fluorouridine triphosphate that mimics UTP, is recognized by RNA 45 polymerase and consequently incorporated into RNA. The most significant action, however, is due to the 5-FU conversion into 5-fluoro-20-deoxyuridine (FdUMP) 2, a known inhibitor of thy- midylate synthetase (TS), a key enzyme in the DNA synthesis.6,7 TS, in the presence of methylene tetrahydrofolate and deoxyuridine monophosphate (dUMP) 3 forms a ternary complex that catalyzes the substitution of 5-H uracil with a methyl group, affording thymine. If FdUMP 50 2 is present, the above ternary complex is not able to carry out this reaction, due to the presence of fluorine in the 5-position: the formation of TMP 4 (20-deoxythymidine 50-monophosphate), the only nucleotide precursor specific to DNA, is, therefore, blocked (Scheme 1), decreasing the availability of TTP (20-deoxythymidine 50-triphosphate) for DNA synthesis.

Scheme 1 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 3

Later two additional mechanisms were proposed for 5-FU 1 antitumor activity: the 55 incorporation of 5-FU into DNA and the alteration of the membrane function of 5-FU 1 treated cells. Recent studies indicate that the TS-direct mechanism predominates when 1 is administered at low doses for a prolonged time, whereas the RNA-mediated process is more active following a bolus administration.8–10 Synthesis of 5-FU 1, by construction of the pyrimidine ring, was first realized by Hei- 60 delberger et al. in 19574,11,12 by reaction of a thiourea derivative 5 with the enolate of ethyl a-fluoro, a-formyl acetate 6 which, in turn, was obtained from methyl formate and ethyl fluoroacetate (Scheme 2). Depending on the chosen a-fluoro-b-ketoester the method is also applicable to the synthesis of other 5-fluoropyrimidines.

Scheme 2 An alternative method for the total synthesis is the direct fluorination of the pyrimi- 13 65 dine ring of compound 7 by means of trifluoromethyl hypofluorite (CF3OF) proposed by Robins in 1971. In the case of CF3OF in methanol/fluorotrichloromethane an intermediate was formed that, after treatment with triethylamine, afforded 5-FU 1 in 84% yield.13 If the reaction was carried out with CF3OF in trifluoroacetic acid 1 was directly isolated in 85% yield.14 The method proposed by Robins13 is suitable also for the direct introduction 70 of fluorine on preformed nucleosides, for example compound 8 (Scheme 3).

Scheme 3 4 Ferraboschi, Ciceri, and Grisenti

The mechanism of the reaction with CF3OF in methanol followed by treatment with triethylamine was later investigated by Robins et al.15 who assigned the structure of (§)-cis-5-ﬂuoro-6-methoxy-5,6-dihydrouracil to intermediate 9 that, by treatment with triethylamine, afforded 5-FU 1 (Scheme 4).

Scheme 4

85 The use of fluorine as fluorinating agent16,17 requires efficient dissipation of the heat of reaction, in order to avoid the destruction of the carbon skeleton of pyrimidine. This result can be achieved by bubbling a mixture of fluorine and an inert gas through a cold liquid, 90 or removing the heat of reaction by carrying out the reaction in the presence of a metal packing or, finally, by addition of very large amount of an inert diluent gas. This last procedure is the most followed and usually fluorine is diluted with an equal amount of nitrogen and then passed through the reaction mixture. 5-FU 1 was obtained in 92.4% yield16 and sublimation at 190C and 1 mm Hg provided a highly pure product.17 95 Xenon difluoride can be used for direct fluorination of the pyrimidine ring but it is diffi- cult to handle due to its high reactivity; in 1980 Kagan et al. realized the direct fluorination 18 of uracil 7 employing C19XeF6, which is much more stable than free xenon hexafluoride. Uracil 7 was the starting material of the above described direct fluorination methods.13–18 In a method patented in 1979,19 cytosine 10 was fluorinated by means of fluorine fluorosulfo- 100 nate (FOSO2F) diluted with nitrogen (60%) affording 5-FU 1 (87.7% yield) (Scheme 5).

Scheme 5 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 5

Orotic acid20 11 was, instead, the starting material for a synthesis of 5-FU by a combination of a ﬂuorination and a decarboxylation (Scheme 6). In the course of the reaction an intermediate was formed that can be converted into 5-ﬂuoro orotic acid 12 (in boiling water) which was, in turn, transformed into 5-FU by heating at 240C.

Scheme 6

105 Starting from the methyl ester of uracil 5-carboxylic acid 13 the synthesis of 5-FU 1 was realized in one-pot, in excellent yield (92%), by addition of a reducing agent (sodium bisulﬁte) aimed to exclude the formation of peroxides by reaction of water with ﬂuorine (Scheme 7).21,22

Scheme 7 A different approach to the synthesis of 1 provided the substitution of the halo atoms of the 2,4,5-trichloropyrimidine or 2,4-dichloro,5-bromopyrimidine 14 by means of potassium ﬂuo- 110 ride at 400C to afford the corresponding 2,4,5-triﬂuoropyrimidine 15; its treatment with sodium hydroxide in water at 80C gave the desired 5-FU 1.23 The starting 2,4,5-trihalopyrimi- dine 14 was obtained from uracil 7 by reaction with chlorine, or bromine, followed by treatment of the 5-halouracil 16 with phosphoryl chloride (76 or 82%, respectively) (Scheme 8).

Scheme 8 6 Ferraboschi, Ciceri, and Grisenti

Some years later a similar approach24 starting from tetrafluoropyrimi- 115 dine25 17, was described by Baasner et al. 4,6-Dichloro-2,5-difluoropyrimidine 18 was obtained in 64% yield from tetrafluoropyrimidine 17 by reaction with gaseous hydrogen chloride. Chlorine in the 6-position was selectively removed by hydrogenation in the presence of palladium on carbon in ethyl acetate and triethylamine (70.5% yield). The resulting 4-chloro-2,5-difluoro- 120 pyrimidine 19 was hydrolyzed with sodium hydroxide in water affording 5-FU 1 in 93% yield. Direct fluorination of uracil continued to be of interest. In 1981 a method based on the use of fluorine/nitrogen in acetic acid was proposed: the intermediate 6-acetoxy-5-fluoro-5,6-dihydrouracil 20, obtained in 70% yield, was converted into the desired 5-FU 1 125 by acetic acid elimination.26 (Scheme 9).

Scheme 9

In a similar approach the use of fluorine diluted with helium in aqueous phosphoric acid27 (Scheme 9)afforded5-FU1 in 84.5%. The advantages of this method are the low cost of phosphoric acid, the capability of its solution to work as an adsorbent of the by-product hydrogen fluoride and to precipitate the final product as 130 crystals. Another fluorinating agent, acetyl hypofluorite, (generated in situ from fluorine and acetic acid) was studied, in comparison with fluorine in acetic acid, from a mechanistic point of view.28 The two geometric isomers of 5-fluoro-6-ace- toxyuracil 20 were isolated as reaction intermediates; their configuration cannot 135 be assigned by 1Hand19F NMR spectra since the coupling constants of cis and trans compounds are too similar. Only one of the isomers, by addition of acetate ion, was transformed into 5-FU 1, acetate being a strong enough base to facilitate acetic acid elimination. The cis and trans configurations were assigned consider- ingthatinthecis isomer the H atom and the acetate group are in trans position, 140 which is the favored one for the elimination of acetic acid. In presence of triethylamine both isomers afforded 5-FU 1 (Scheme 10). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 7

Scheme 10

Visser et al.29–30 also studied the mechanism and the stereochemistry of uracil 7 fluorination 145 with fluorine or acetyl hypofluorite, proposing that addition at the 5–6 double bond occurs through radical anion 21; formed 5-fluoro-6-acetoxy derivative 20 evolves into the final product through an intermediate that in the presence of a base affords final 5-FU 129 (Scheme 11). The influence of the N-1 substituent on the stability of 5-fluoro-6-acetoxy intermedi- ate31 and conversion by reaction with alcohols were also studied.30

Scheme 11

150 A commercially available, stable, easy to handle ﬂuorinating agent, SELECT- FLUOR, was proposed in 1995 for the synthesis of ﬂuorohydrin 22 that, by sublimation, provided 5-FU 1 (82%)32 (Scheme 12). 8 Ferraboschi, Ciceri, and Grisenti

Scheme 12 Some approaches avoiding the direct ﬂuorination are reported in the literature. For example, starting from s-triazine 23, it is possible to prepare 5-FU 1 (88% yield) by reac- 155 tion of ﬂuoroacetamide and lithium diisopropylamide (Scheme 13).33

Scheme 13 Diethyl fluoromalonate (prepared from trifluoroacrylic acid) was easily converted to 5-fluoro-6-chlorouracil 25 by reaction with urea to afford 5-fluorobarbituric acid 24 (97% yield) followed by reaction with phosphoryl chloride in dimethylaniline (46% yield). The 6-chlorine atom of 25 was removed by hydrogenolysis (Pd/C) affording 5-FU 1 in 73– 160 91% yield, depending on the reaction conditions (Scheme 14).34

Scheme 14 The methyl ester of 2-ﬂuoro-3-methoxy acrylic acid 26 can be used as a synthon of the ﬂuorinated uracil ring: by reaction with O-(trimethylsilyl)urea intermediate 27 was formed that by cyclization with sodium hydroxide afforded 5-FU 1 (67% yield) (Scheme 15).35

Scheme 15 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 9

Recently,36 5-FU 1 was prepared by deamination of 5-fluorocytosine 28 by means of 165 a new 5-methylcytosine deaminase: expression of deaminase in E. coli caused an efficient transformation of non-toxic 5-fluorocytosine 28 into 5-FU 1 that abolished or severely inhibited growth of cells. The goal of this study was, of course, to transfect cancer cells with the deaminase (Scheme 16).

Scheme 16

5-FU 1, since its introduction more than 50 years ago, has become a component of 170 therapy for gastrointestinal, head and neck and breast cancers. The activity of 5-FU is limited by its rapid degradation into 5,6-dihydro-5-ﬂuorouracil (5-FUH2) 29 under the action of dihydropyrimidine dehydrogenase (DPD), an enzyme NADPH-dependent (Scheme 17).5 It has been demonstrated that this enzyme deactivates more than 85% of the injected 5-FU 1.

Scheme 17

175 At the beginning 5-FU 1 was administered by intravenous bolus but its low biological t1/2 makes continuous infusion more convenient. The use of oral 5-FU 1 was abandoned some decades ago because of its irregular absorption, due to intra- and inter-individual differences depending on the variable activity of dihydropyrimidine dehydrogenase. In addition to the inconvenience of i.v. administration, the efficiency 180 of 5-FU 1 is limited by its toxicity due to phosphorylation in the digestive tract and to the lack of selectivity toward tumors. The development of 5-FU 1 derivatives was the target of many studies. The toxicity of 5-FU 1 can be reduced by derivatives which are stable to enzymatic degradation, by derivatives that inhibit DPD or by prodrugs of 1 that liberate the active principle in tumor cells. A prodrug is defined 185 as a pharmacologically inactive compound that is converted into an active agent by metabolic transformations. The prodrugs of 5-FU 1 are characterized by a pyrimidine ring bearing a fluorine atom in the 5 position. The main benefit is oral administration, with the improvement in quality of life of the patient. 10 Ferraboschi, Ciceri, and Grisenti

In the present review we now describe some derivatives in which 5-FU 1 is conju- 190 gated to drug carriers or molecules endowed with antitumor activity.

2. 5-Fluorouracil Derivatives In 1985 N1-(3-aminopropyl)-5-fluoro uracil 30 was synthesized from 5-FU 1 by treatment with hexamethyldisilazane and trimethylchlorosilane followed by reaction with N-(3-bro- mopropyl)-phthalimide. Deprotection afforded the desired product in 54% yield. The 195 amino derivative was considered a suitable precursor of N1-(2-formylethyl)-5-fluorouracil 31, which may be converted to 5-FU following an enzymatic reaction. Indeed, in presence of amine oxidase the amine was transformed into 5-FU 1 and acrolein, in the course of 24 h incubation. The slow enzymatic conversion, however, indicated that 5-FU-derivative 30 might not be efficiently metabolized in vivo (Scheme 18).37

Scheme 18

200 A mutual prodrug consists of two synergistic drugs joined together. The two molecules may be connected directly or by means of a linker: in 1997 work38 5-FU 1 was linked to cytarabine 32, an antitumor agent used mainly in the treatment of acute leukemia and lymphomas, affording compound 33. Cytarabine (ara-C) 32 is attached to the double drug through a hydrolyzable amide bond, while 5-FU is attached via an acyloxy- 205 methylene group, easily removable. Two spacers with different length (two or four methylene groups) were used and the synthesis is outlined in Scheme 19. The synthesis afforded the product with yield ranging from moderate to excellent; and, after the introduction of 5-FU 1, mild conditions and selective reagents, due to the formation of the labile N-O ketal derivative, were required. The benzenesulfenyl chloride, by-product of 210 the preparation of chloromethylester 34, was removed by trapping it with cyclohexene- 1,2-dicarboxylic acid, instead of the usual cyclohexene, in order to remove the addition product by an aqueous washing. This avoided separation by distillation of products hav- ing similar boiling points. Final compound 33 with n D 4 was stabler than the corresponding compound with n D 2, over a wide range of pH, and for this reason it was estimated to 215 be more suitable for biological testing. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 11

Scheme 19

More recently39 another mutual prodrug of 5-FU 1 was developed with diazenium- diolates, which are known to be controlled sources of NO. NO can inhibit metastasis, enhance cancer cells apoptosis and assist macrophages to kill tumor cells. Two different conjugates, with methylene or acyloxymethylene as spacers, were synthesized starting 220 from sodium 1-(pyrrolidin-1-yl)diazen-1-ium-1,2-diolate (PYRRO/NO) 35 in turn prepared according to a procedure already reported.40 The key intermediate of the two syntheses was the O2-chloromethyl 1-(pyrrolidin-1-yl)diazen-1-ium-1,2-diolate (chloromethyl PYRRO/NO) 36 that by direct reaction with 5-FU 1 or through its succinic acid derivative 37 afforded the desired products 38 together with the N1,N5-bisalkylation prod- 225 ucts 39 (Scheme 20). The two mono-alkylated conjugates were evaluated in hydrolysis by monitoring the NO release. Both prodrugs showed greater activity than 5-FU, but the 12 Ferraboschi, Ciceri, and Grisenti

Scheme 20

compound with methylene as spacer was very stable in aqueous solution and no substan- tial NO was detected. On the contrary the other prodrug with a longer spacer released NO both at pH 8 and in the presence of esterase. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 13

230 In order to enhance the delivery characteristics of 5-FU 1 a series of N-acyl and N-alkoxycarbonyl derivatives were prepared and their antitumor activity evaluated.41,42 According to a 1980 article41, the most promising antitumor agent, even when administered orally, was the N1-acetyl-N3-ortho-toluyl-5-ﬂuorouracil 40. The paper 235 described the different acylated products that can be obtained depending on the chosen acylating agent, solvent and reaction temperature. N1-Acetyl-N3-ortho- toluyl-5-FU 40 was prepared starting from N1-acetyl derivative 41 by reaction with ortho-toluyl chloride in dioxane and triethylamine. The N3-ortho-toluyl derivative of 5-FU, 42, was isolated in mice serum when N1-acetyl-N3-ortho-toluyl-5-FU 40 240 was orally administered. In order to obtain this N3-monosubstituted compound 42 the best conditions used ortho-toluyl chloride in pyridine at room temperature (Scheme 21).

Scheme 21

Further, the N1-alkoxycarbonyl derivatives42 prepared from 5-FU 1 by reaction with the appropriate chloroformates in pyridine were hydrolyzed in serum affording 245 the 5-FU 1. b-Lactam based prodrugs, in particular cephalosporin-based prodrugs, have been reported in the past and in 200943 this approach was also applied to 5-FU 1. The 5-FU-cephalosporin conjugate was synthesized starting from the N1-BOC-5-FU 43 and the suitably functionalized cephem 44 (Scheme 22), 250 affording a compound stable in aqueous media; conjugate 45 in the presence of b-lactamase was completely cleaved to 5-FU 1 and the hydrolyzed cephalosporin, demonstrating its strategic potential against a range of human carcinoma cells. The observation that the phthalide fragment was widely used in the creation 255 of transport forms of antibiotics prompted the introduction of this moiety in the 14 Ferraboschi, Ciceri, and Grisenti

Scheme 22

5-FU 1 molecule.44,45 1-(3-Phthalidyl)-5-fluorouracils were synthesized starting from the N3-benzoyl-5-FU 4644 or from the bis(trimethylsilyl) derivative of 5- FU 4745 (Scheme 23). In the first case N3-benzoyl-5-FU 4644 was reacted with 3-bromophthalide in presence of sodium hydride affording phthalide derivative 260 48 in 59% yield. Subsequent acidic hydrolysis gave desired product 49 in 54% yield. The phthalide derivative was effectively biotransformed into 5-FU 1 in mice serum. Starting from 2,4-bis(trimethylsylil)-5-fluorouracil 47 45 phthalide derivative 49 was obtained in 94% yield by reaction with 3-bromo phthalide at 100C. 265 Some 5-FU prodrugs with a removable attachment were also proposed. For example porphyran 50, a polysaccharide from the red algae Porphyra haitanensis, was employed as a drug carrier, allowing the fixation at 5-FU 1 at the 6-position through an acetyl spacer group of polysaccharide derivative 51 (Scheme 24).46 The release of 5-FU 1 from conjugate 52 was studied in 0.01M NaOH solution, in 0.1M HCl solution and in phosphate 270 buffer. The amount of released 5-FU 1 was significantly different, and only in basic media did the prodrug release quickly the 5-FU 1. In vivo studies are not reported. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 15

Scheme 23

Scheme 24

In order to decrease the toxicity of 5-FU 1,2,5-triacetyl-b-D-glucofuranurono-6,3- lactone 53 was reacted with 2,4-bis(trimethylsylil)-5-ﬂuorouracil 47, according to a reported method.47 The obtained molecule combines the active principle 5-FU 1 and the 275 nontoxic glucuronic acid, which participates in the detoxiﬁcation of xenobiotics in the organism by forming the glucuronides (Scheme 25).48 Biological studies showed that 16 Ferraboschi, Ciceri, and Grisenti

Scheme 25

5-FU derivative 54 has a very low toxicity and a relatively high antitumor activity. At pH 1.4 only the lactone ring undergoes splitting while at pH 7.4 the lactone ring and the N-glycosidic bond are split with formation of 5-FU 1. 280 The concept of “radioinduced drug” proposed for cancer therapy is also noteworthy: 5-FU derivatives variously substituted at the 1-position were prepared and the release of the antitumor drug was observed upon g-irradiation.49 5-FU dimer 5550 and 5-ﬂuoro-1-(20-oxocycloalkyl)- uracil 5651 released 5-FU 1 by radiation activation under hypoxic conditions (Scheme 26).

Scheme 26

5-FU 1 can also be released from gold nanoparticles by photoirradiation.52 In this 285 case 1 was conjugated to gold nanoparticles through a photoresponsive ortho-nitrobenzyl linkage. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 17

Photoirradiation was also responsible for the recently reported53 release of 5-FU from oligonucleotide tetramer 58 (Scheme 27). The key reaction was the elimination of a nucleobase by photoinduced C10 hydrogen abstraction, in turn caused by the radical (compound 290 59) photo-generated at C5 of the adjacent 5-halouracil. This approach seems to be promising, since the 5-FU could be released by photoirradiation only in the irradiated area of the body.

Scheme 27

Even if they are certainly promising, the 5-FU derivatives described herein have not yet found application in clinical trials, although some of them are under investigation, and studies to assess their possible therapeutic use are still in progress.

295 II. Fluoropyrimidine Nucleosides Among the 5-FU derivatives developed for oral administration, 5-ﬂuoropyrimidine nucleosides deserve an outstanding position as proven by a 2000 review5 about the 5-FU prodrugs considered from the clinical point of view. Protracted oral administration should be the ideal route of administration of 5-FU, being preferred by the patients and by pharmacoeconomic considera- 300 tions. The low oral bioavailability of 5-FU 1 has been overcome by the new generation of oral ﬂuoropyrimidine nucleosides.54–56 The main 5-FU prodrugs are depicted in Scheme 28. 18 Ferraboschi, Ciceri, and Grisenti

Scheme 28

1. Tegafur (Ftorafur) Tegafur 60, also known as ftorafur (1-(tetrahydro-furanyl)-5-fluorouracil) is the first designed 5-FU prodrug. It has a high chemotherapeutic activity (twice that of 5-FU) and 305 low toxicity (5–6 times less than 5-FU) and it is used for the treatment of breast and gastrointestinal tract cancer. Typically, it is used in combinations with other drugs to further improve its bioavailability.54–57 For instance, the combination of uracil 7 and tegafur 60 (4:1), namely UFT, allows for higher levels of circulating 5-FU, by saturating the dihydropyrimidine dehydrogenase with its natural substrate uracil 7. 310 Tegafur was synthesized for the first time in 196958 through the condensation of the bis(trimethylsilyl) derivative of 5-FU 47 and 2-chlorotetrahydrofuran 64 (Scheme 29)or by condensation of 5-fluorouracil-mercury and 2-chlorotetrahydrofuran 64 in dimethylformamide and toluene (75%).

Scheme 29

2-Chlorotetrahydrofuran 64 was also used by other authors59–61, and its preparation 315 was the object of studies60 that included determination of the best molar ratios61 between the reagents in order to minimize the formation of the 1,3-bis(tetrahydro-20-furyl)-5-ﬂuo- rouracil that is the main by-product due to the double condensation reaction. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 19

Tegafur 60 is administered as a racemic mixture, due to no signiﬁcant differences in the effects of the stereoisomers, as conﬁrmed by the tests carried out with the separated 320 enantiomers in 1977.62 The separation of enantiomers was achieved by formation of diastereoisomers with brucine (Scheme 30). The racemic mixture of tegafur was, in this case, prepared in 91% yield, by reaction of 2-acetoxy tetrahydrofuran 65 with the silylderivative of 5-FU 47, using sodium iodide as catalyst.

Scheme 30

The same 2-acetoxytetrahydrofuran 65 was also used with typical Friedel-Craft cata- 325 lysts63 or without catalyst at 120C in dimethylformamide (Scheme 31); higher temperatures lead to the degradation either of tegafur 60 or of 2-acetoxytetrahydrofuran 65.64

Scheme 31

A different catalyst, cesium chloride, was used in the condensation between disily- lated 5-FU 47 and 2-acetoxytetrahydrofuran 65, in acetonitrile affording tegafur 60 in 87% yield (Scheme 32).65 20 Ferraboschi, Ciceri, and Grisenti

Scheme 32

330 The most popular tetrahydrofuran derivative in the synthesis of tegafur 60 was, however, the 2,3-dihydrofuran 66. In a French patent, an excess of this compound was reacted with 5-FU 1 in aprotic polar solvents affording tegafur 6066 (Scheme 33) with yield depending on the chosen solvent.

Scheme 33

In similar conditions but in the presence of an acidic catalyst (for example triethyl- 335 amine hydrochloride, and tetrabutoxytitanium) 59–89% yields were observed.67 The same reaction was realized in the presence of phosphorus pentachloride in hexamethyl- phosphoramide affording tegafur 60 in 88% yield; poor yields were observed changing the solvent (DMF or DMA) and a decreased regioselectivity was obtained using phosphorus trichloride instead of phosphorus pentachloride.68 340 Two patents of two different Japanese companies described the synthesis of tegafur 60 from 5-FU 1 and 2,3-dihydrofuran 66 in the presence of trimethylsilyl chloride and triethylamine in dimethylformamide69 or in the presence of dimethyldichlorosilane and triethylamine in acetonitrile.70 The optimization of the molar ratio between 5-FU 1 and 2,3-dihydrofuran 66 and the use of calcium chloride as catalyst, under pressure, afforded 345 very high yields (85–92%) of tegafur 60.71 A one-pot synthesis72 was realized starting from 2,3-dihydrofuran 66 that was treated with water in the presence of an acidic catalyst (H3PO4,H2SO4, alkylsulfonic acids) at a pH value below 2.5; after removal of the excess of unreacted material the obtained mixture was treated with the silyl derivative of 5-FU 47 and Friedel-Craft catalyst (SnCl4, 350 BF3, TiCl4, NaI) (Scheme 34). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 21

Scheme 34

Depending on the ratio between 2,3-dihydrofuran 66 and 5-FU 1, in pyridine, in a sealed tube, tegafur 60 or 1,3-bis(tetrahydrofuryl)-5-ﬂuorouracil 67 were obtained (Scheme 35).73 Interestingly, the heating of the bis-substituted 5- FU 67 at 135C for 12 h, at reduced pressure, afforded tegafur in 76% yield.

Scheme 35

355 Some different starting materials were also proposed, alternative to the 5-FU or to the tetrahydrofuran derivatives. For example, the 5-ﬂuoro-6-hydroxy-5,6-dihydrouracil 6874 provided, by reaction with hexamethyldisilazane or hexamethyldisilazane and trimethylsilyl chloride, bis(trimethylsilyl)-5-ﬂuorouracil 47 (80–82%), whichbyfurtherreactionwith 2-chlorotetrahydrofuran 64, in THF, afforded tegafur 360 60 (51% yield) (Scheme 36). 22 Ferraboschi, Ciceri, and Grisenti

Scheme 36

A series of 2-alkoxyderivatives was used, the best results being achieved with the 2-tert-butoxytetrahydrofuran 69 (67% yield).75 The small amount of disubstituted by-product 67 was quantitatively converted into tegafur by treatment with acetic acid (Scheme 37).

Scheme 37

365 In a different approach, a preformed nucleoside, b-5-ﬂuoro-20-deoxy-uridine 61, was used as starting material and submitted to the reaction with black platinum in water and sodium hydrogen carbonate; the recovered 50-carboxyderivative 70 was transformed into the corresponding methyl ester 71 and the 30-hydroxy group was derivatized as mesylate 0 0 72. Simultaneous elimination of 3 -O-mesylate and 5 -COOCH3 afforded the dihydro- 370 furan ring (compound 73); catalytic hydrogenation gave the (20R) isomer of tegafur 60 (Scheme 38).76 Starting from the a-preformed nucleoside the (20S) enantiomer of tegafur 60 was obtained. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 23

Scheme 38

The dimethylacetal of 4-hydroxybutanal-4-O-trimethylsilylether 74 was used as precursor of the tetrahydrofuran ring:77 the acetal was treated with bis(trimethylsilyl)-5-ﬂuo- 375 rouracil 47, in acetonitrile, in the presence of stannic chloride. Tegafur 60 was recovered in 84% yield (Scheme 39).

Scheme 39

bis(Trialkylstannyl)-5-ﬂuoro uracil 7578 was used, instead of the usual silyl derivative, in the reaction with a series of 2-substituted tetrahydrofurans, in the presence of 24 Ferraboschi, Ciceri, and Grisenti

different acid catalysts. The formation of desired tegafur 60 (70–95%) was obtained when 380 a 1:1 molar ratio was utilized of 5-FU derivative/2-substituted tetrahydrofuran. In the case of a molar ratio 1:2 the N1,N3-disubstituted product was isolated. As previously reported75 N1,N3-disubstituted product 67 was converted into tegafur 60 by acidic treatment (Scheme 40).

Scheme 40 As in the case of 5-FU 1 synthesis the direct fluorination of the preformed nucleoside 385 was described. Trifluoromethyl hypofluorite in dichloromethane79 gave the fluorination of 1-(tetrahydrofuranyl) uracil 76 in 60% yield (Scheme 41).

Scheme 41 Using fluorine as fluorinating agent, the reaction with 1-(20-tetrahydrofuryl)-5-meth- oxycarbonyluracil 77 gave different intermediates depending on the chosen solvent.21,22 In water 5-fluoro-6-hydroxy intermediate 78 was obtained, while in acetic acid 5-fluoro-6-ace- 390 toxy derivative 79 was formed. Both intermediates led to desired tegafur 60 in 84% and 62% yield, respectively (Scheme 42)byhydrolysisofthemethylester and elimination reaction. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 25

Scheme 42

More recent works about tegafur are especially focused on the improvement of its oral absorption and biological half-life and to lowering its toxicity. With this aim in mind, a light- triggered porphyrin tegafur prodrug was developed.80 Porphyrins are photosensitizers used in 395 photodynamic therapy and they tend to accumulate in neoplastic tissue to higher concentrations than in adjacent normal tissue. The porphyrin prodrug was composed of three parts: a

Scheme 43 26 Ferraboschi, Ciceri, and Grisenti

porphyrin, a photocleavable ortho-nitro benzyl moiety, as a light-triggered group, and tegafur. In vitro tests demonstrated that this prodrug is less toxic than its parent drug and that it can release cytotoxic tegafur 60 upon photoactivation with long wavelength UV light (350 400 nm). The synthesis of porphyrinic prodrug 80 is depicted in Scheme 43.

2. Floxuridine Floxuridine 61 (5-fluoro-20-deoxyuridine, FdUrd) (See Scheme 28) is the deoxyribose metabolite of 5-FU and the precursor of 5-fluorouridine monophosphate 2 (FdUMP) that inhibits thymidilate synthetase (TS) (See Scheme 1). FdUrd 61 can be also converted into 405 5-FU 1 in the liver by thymidine phosphorylase (TP). It was approved in therapy by the FDA in 1970 and has been extensively used for the clinical treatment of carcinoma of the ovary, breast and gastrointestinal tract. Due to the higher toxicity, costs and equal efficacy compared to 5-FU 1 the use of FdUrd 61 is restricted to the treatment of liver metastases caused by colorectal cancer.81 It was first synthesized at Hoffman-La Roche through the 410 reaction (Scheme 44)82,83 of a mercury derivative of 5-FU 81 with a protected 2-deoxy- D-ribosyl chloride 82 (68–70% yield) followed by separation of a and b-anomers by crystallization (42.5% of b-isomer) and removal of protecting groups (90%). The same company obtained FdUrd 61 and its phosphates through a biocatalytic approach (cells of Streptococcus faecalis).84

Scheme 44 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 27

415 A subsequent approach provided the formation of 20-deoxyribose starting from a preformed b-nucleoside, namely 5-ﬂuorouridine 83. This purpose was achieved by means of the removal of 20-chlorine 8585 or 20-bromine 8686 in turn obtained by the ring opening of aO2,2-anhydrouridine 84 (Scheme 45). Use of tributyltin hydride allowed removal of the 20-halogen selectively.

Scheme 45 28 Ferraboschi, Ciceri, and Grisenti

420 Applying a method developed in 197487 for the synthesis of 20-deoxyuridine from uridine, ﬂoxuridine 61 was obtained in 50% yield from 5-ﬂuorouridine 83. The 20- hydroxy group was substituted by a bromine atom that was removed by hydrogenation (Scheme 46).88 A mechanism of the substitution reaction was proposed.

Scheme 46

Similar to 5-FU 1 and tegafur 60, the direct fluorination of the peracetylated pre- 425 formed 20-deoxy nucleoside was realized using, as fluorinating agent, trifluoromethyl hypophosphite.13,15 Perbenzoate 20-deoxyuridine 87 was fluorinated with fluorine in acetic acid (60–90% yield). The benzoyl esters were removed from compound 88 by metha- nolysis affording floxuridine 61 (60–90% yield)89 (Scheme 47).

Scheme 47 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 29

In a different approach, a study, carried out with the aid of 1H NMR, allowed workers 430 to establish which parameters inﬂuence the stereoselectivity of the reaction between the bis-silyl derivative of 5-FU 47 and 3,5-bis(O-para-chlorobenzoyl)-2-deoxy-a-D-ribofura- nosyl chloride 89. In the presence of para-nitrophenol the b-anomer of 90 was obtained stereoselectively in high yields. In the presence of pyridine and para-nitrophenol, the a-anomer of 90 was formed stereoselectively (Scheme 48).90

Scheme 48 435 A biocatalytic approach to floxuridine 61 includes the use either of purified enzymes or of microorganisms in toto. The regioselective behavior of hydrolytic enzymes towards the esters of floxuridine was already observed in 1989. A lipase from Pseudomonas fluo- rescens was able to regioselectively remove the 30-acyl group from the 30,50-diester of floxuridine 91. The selective removal of the 50-acyl group was achieved by means of sub- 440 tilisin, the protease from Bacillus subtilis (Scheme 49).91

Scheme 49 30 Ferraboschi, Ciceri, and Grisenti

A synthesis of ﬂoxuridine 61 starting from 2-deoxyribose 92 was performed in 2008 through a chemo-enzymatic approach: CAL B, the lipase from Candida antarctica, was ﬁrst used to selectively introduce an acetyl group at the 5-position of the carbohydrate (compound 93) and, in the next step of the synthesis, to selectively remove the same acyl 445 group, under alcoholysis conditions (compound 94). Intermediate 5-phosphate 95 was used as starting material for the glycosylation reaction catalyzed by a combination of phosphopentomutase (PPM) and thymidine phosphorylase (TP). PPM catalyzed the trans- fer of the phosphate group from the 5-position to the 1-position of the furanose, providing the substrate of TP. Floxuridine 61 was obtained in 60% yield from 2-deoxy-D-ribose-5- 450 phosphate 95 (Scheme 50).92

Scheme 50

The use of TP as biocatalyst of the nucleobase exchange reaction, aimed to convert thymidine 96 to floxuridine 61 (transglycosilation), was also reported. In one case immobilized TP was used (62% yield);93 by using commercially available TP94 a 86% yield was observed (Scheme 51). 455 Immobilized bacterial cells from E. coli95 or from Lactobacillus animalis 96,97 were reported to catalyze the transglycosilation reaction of thymidine 96 to floxuridine 61with 62% and 95% yield, respectively. Similarly to 5-FU 1, floxuridine 61 has a short plasma half-life and causes gastrointestinal toxicity. Some esters of floxuridine with long chain aliphatic acids show higher 460 therapeutic indices than floxuridine98 and for this reason the susceptibility of a series of Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 31

Scheme 51

30,50-diesters or 30-and 50-monoesters to porcine liver esterase was investigated. The obtained results suggested that the higher antitumor activity of longer alkyl chain diesters of floxuridine was partly due to their slow rate of hydrolysis by non-specific esterase. 465 Many efforts have been made to develop a variety of prodrugs undergoing hydrolysis to release floxuridine 61: for example, conjugates with a cyclic peptide with an ester linker99 or a peptide with a 3-aminoxypropionate-based linker.100 This last prodrug 97 shows a linker that cyclizes readily under physiological conditions and in the interstitial tissues of solid tumors (Scheme 52).

Scheme 52 32 Ferraboschi, Ciceri, and Grisenti

470 Floxuridine binding to antibodies in order to drive the cytotoxic activity101,102 was also studied; and further photolabile carbonate prodrug 98 was synthesized that, by pho- tolysis, rapidly released ﬂoxuridine 61 (Scheme 53).103

Scheme 53

It is known (see Scheme 1) that floxuridine 61 is activated by phosphorylation, catalyzed by thymidine kinase (TK), to 5-fluorodeoxyuridine monophosphate (FdUMP) 2, the TS-inhibitor. 475 In TK-deficient tumors floxuridine lost its cytostatic potential. Direct administration of phosphorylated floxuridine has little therapeutic advantage since the charged monophosphate, under physiological conditions, shows poor, if any, penetration across the cell membrane. The administration of lipophilic phosphoramidate 99 can circumvent these problems (Scheme 54).104

Scheme 54

According to a recent patent, a ﬂoxuridine prodrug was prepared and a biorthogonal cleav- 480 age (under biocompatible conditions) was suggested (Scheme 55).105 The target of the study was the development of a prodrug releasing the active drug by means of a reaction susceptible to occur in a biological environment without interfering with the normal functions of its Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 33

Scheme 55 component. Thus, 3-propargyl-5-ﬂuoruridine 100 was synthesized and the free active drug was generated using biocompatible palladium catalyst, at 37C and pH 7.4; also the by-product 1- 485 hydroxyacetone 101 is biocompatible. Palladium implants, can, therefore, deprotect the prodrug at the disease site, reducing the general systemic concentration of the free drug.

3. Doxifluridine Doxifluridine (50-deoxy-5-fluorouridine, 50-DFUR) 62 is a prodrug that requires thymidine phosphorylase (TP) for its one-step conversion to 5-FU 1. Since TP expression is high in the gas- 490 trointestinal tract, doxifluridine therapy resulted in dose-limiting toxicity, such as diarrhea. It is orally administered for the treatment of breast, stomach, colon cancer and nasopharyngeal carcinoma. A large number of 50-deoxy nucleosides have been synthesized taking into account that several biologically active nucleosides require the presence of a 50-hydroxy group for 495 activation, usually by phosphorylation. Removal of this function provides interesting compounds for biochemical and biological studies. 50-Deoxy compounds are potentially interesting medicinal agents, since such molecules cannot be phosphorylated and incorporated into nucleic acids and thus offer the possibility of reduced toxicity and increased specificity. The 50-deoxy carbohydrates can be synthesized before the formation of the N- 500 glycosidic bond with the 5-fluoropyrimidine or, alternatively, a preformed nucleoside can be used as starting material for the 50-modification. In a 1982 patent,106 the 5-deoxy carbohydrate was prepared, starting from suitably protected D-ribose 102, through 5-bromo derivative 103 (Scheme 56). Removal of the 5- bromine atom was achieved by means of a catalytic hydrogenation affording compound 505 104 (90% yield). Deoxy peracetylated sugar 105 was used in the glycosylation reaction affording diacetate of doxifluridine 106 that was deprotected by means of sodium methyl- ate (97% yield). The same authors, by this method, prepared also the 50-monodeutero and the 50,50-dideutero derivatives and the a-anomer of doxifluridine.107 34 Ferraboschi, Ciceri, and Grisenti

Scheme 56

A modification of the coupling method between the 50-deoxysugar and the 5-fluoro 510 pyrimidine allowed workers to increase the yields and the purity of the final product, at a temperature below 0C.108 In a 1978 article109 the 50-deoxy group was obtained starting from 5-fluorouridine 83 (Scheme 57), through the simultaneous protection of 20- and 30-hydroxy groups as a cyclic

Scheme 57 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 35

sulphinyl derivative, and the substitution of 50-hydroxy group with a chlorine atom (com- 515 pounds 107), by means of thionyl chloride. The chlorine atom was then removed by reaction with tributyltin hydride. Protection of the 20- and 30-hydroxy groups of 5-ﬂuorouridine as the isopropylidene derivative 108, followed by iodination110 afforded suitable 50-iodo intermediate 109 in order to obtain the 50-deoxy 5-ﬂuoronucleoside 62 (Scheme 58).

Scheme 58

520 The same 50-iodo derivative, compound 109, as later reported in an Italian patent,111 was transformed into the 50-deoxy nucleoside by removal of the acetonide, followed by reaction with tributyltin hydride in toluene (Scheme 59).

Scheme 59 36 Ferraboschi, Ciceri, and Grisenti

Another Italian team in 2009112 described a chemo-enzymatic approach to the synthesis of doxiﬂuridine through the regioselective enzymatic hydrolysis of triacetyluridine 525 110 followed by a phosphorylase-catalyzed transglycosylation (Scheme 60). The regioselective hydrolysis of 50-acetate was achieved by means of lipase from Pseudomonas ﬂuo- rescens in pH 7 buffer; the 50-hydroxy group of compound 111 was transformed into 50- bromo derivative 112 in two steps. Reduction of 50-bromonucleoside (tributyltin hydride) and removal of protecting groups afforded 50-deoxy uridine 113, a suitable substrate for a 530 phosphorylase-catalyzed substitution of the nucleobase.

Scheme 60

4. Capecitabine Capecitabine (N4-pentyloxycarbonyl-50-deoxy-5-ﬂuorocytidine, XelodaÒ) 63, an oral prodrug of 5-FU, was designed to circumvent the gastrointestinal toxicity of doxiﬂuridine 62 and to generate 5-FU preferentially at the tumor site.113–116 The activation of capecita- 535 bine 63 by transformation into 5-FU required three distinct enzyme-catalyzed steps. After Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 37

oral administration capecitabine passes the intestine unaltered; in the liver the carbamoyl moiety is removed by carboxylesterase; the second step is the conversion to 50-deoxy-5- ﬂuorouridine 62 by cytidine deaminase, an ubiquitous enzyme,117 responsible for the deamination of nucleosidic analogs, localized in the liver and in various tumor types. 540 Finally, 50-deoxy-5-ﬂuorouridine 62 can be converted to 5-FU in the tumors by thymidine phosphorylase,118 an enzyme that leads to the preferential generation of 5-FU in tumors as a result of its overexpression in malignant tissue (Scheme 61).

Scheme 61 Capecitabine 63 is approved in the European Union, USA, Canada and Australia for the treatment of advanced or metastatic colorectal carcinoma.119–122 In combination with 545 other antineoplastic agents, it can be utilized also in metastatic breast cancer.123–127 Syn- theses of capecitabine were reviewed in 2010 in a book chapter.128 In the molecule of capecitabine 63 three moieties can be recognized: the 50-deoxy-D-ribose, the 5-ﬂuoropyrimidine and the pentyloxycarbonyl group. Similarly to the doxiﬂuridine synthesis, the N-glycosylation can be carried out starting from D-ribose, postponing the formation of the 50- 550 deoxy group to a subsequent step, or from the previously prepared 50-deoxy-D-ribose. Also the N4-carbamoyl function can be introduced in the course of different steps of the synthesis. Several syntheses start from triacetyl-5-deoxy-D-ribose 105 that, in turn, can be prepared from methyl 2,3-O-isopropylidene-D-ribofuranoside 115 through the formation of 5-O-mesyl derivative 116, the substitution of the mesylate group by an iodine atom (compound 117)and 555 the reductive removal of the latter. Deprotection and acetylation followed by crystallization afforded the b-anomer of 1,2,3-O-triacetyl-5-deoxy-D-ribose 105 (Scheme 62).129

Scheme 62 38 Ferraboschi, Ciceri, and Grisenti

Peracetylated 5-deoxy-D-ribofuranoside 105 was also prepared in a different way, starting from 2,3-O-isopropylidene-D-ribofuranoside 115, through the 5-bromo-derivative and used in the N-glycosylation step, as described in the case of doxiﬂuridine synthe- 560 sis (see Scheme 56).106 In more recent work130 the removal of the 5-hydroxy group was realized through the reduction of suitably protected 5-O-tosyl ester 119 by means of lithium aluminum hydride (Scheme 63); after deprotection and acetylation the 1,2,3-O-triacetyl-5-deoxy-D-ribose 105 was obtained as a 1:3 a/b anomeric mixture. A ratio more favorable to the b-anomer 565 was observed131 by modiﬁcation of the reaction conditions of 5-deoxy-D-ribose acetylation; indeed, using triethylamine in dichloromethane solution as the base, instead of pyridine, the acetylation afforded a 1:13.8 a/b anomeric mixture (Scheme 63). Pure b-isomer 105 was recovered in 80% overall yield by crystallization.

Scheme 63

A different reducing agent for 119 (sodium borohydride in dimethylsulfoxide)132 570 allowed the synthesis of 1,2,3-O-triacetyl-5-deoxy-D-ribose 105 in 56% overall yield from D-ribose (Scheme 64). Pure b-anomer was recovered after crystallization (60%).

Scheme 64 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 39

An improvement of this method was described in a 2013 Indian patent:133 5-O-Tosyl derivative 119 was obtained, avoiding the use of pyridine, by reaction of 1-O-methyl-2,3- O-isopropylidene-D-ribose 115 with para-toluensulfonyl chloride and sodium hydroxide 575 in toluene, in the presence of tetrabutyl ammonium bromide (Scheme 65).

Scheme 65 In 2000,134 1,2,3-O-triacetyl-5-deoxy-D-ribose 105 was used as starting material for the glycosylation of the silyl derivative of 5-ﬂuorocytosine 120 in the presence of stannic tetrachloride or trimethylsilyl iodide. Reaction with n-pentyl chloroformate and removal of protecting groups afforded the desired capecitabine 63 (Scheme 66). Instead of stannic 580 tetrachloride, triﬂic acid can be used as acid catalyst.135

Scheme 66 The N-glycosylation was also carried out without silylating agent, in a “one-pot” procedure that afforded capecitabine 63 in 60% yield (Scheme 67).136

Scheme 67 40 Ferraboschi, Ciceri, and Grisenti

Other syntheses were designed starting from a differently protected 5-deoxy-D- ribose. For example 5-deoxy-D-ribose 118 was transformed into corresponding ace- 585 tonide 121 by reaction with 2,2-dimethoxypropane in the presence of para-toluen- sulfonic acid, in dimethylformamide as solvent. Acetylation of the 1-hydroxy group afforded the suitable intermediate, compound 122, for the glycosylation of N4-pentyloxycarbonyl-5-ﬂuorocytosine 123 under the usual conditions. Acidic treatment (Amberlyst 15) of obtained 124 afforded ﬁnal capecitabine 63 590 (Scheme 68).137

Scheme 68

The 1,2,3-hydroxy groups of 5-deoxy-D-ribose 118 were also protected as methyl carbonates (compound 125) providing a 1:2.7 a/b anomeric mixture (63%) that was directly used in the glycosylation step. 20,30-Protected-50-deoxy-5-fluorocytidine 126 was recovered, after precipitation of undesired by-products, and submitted to the reaction 595 with pentyl chloroformate; removal of 20,30-carbonates under basic conditions and crystallization afforded capecitabine 63 (Scheme 69).138 A mixture of a- and b-anomers of 1,2,3-tri-O-acetyl-D-5-deoxy-ribose 105, was also used in the glycosylation of N4-pentyloxycarbonyl-5-fluorocytosine 123, that is on the 5- 139 fluorocytosine already derivatized on the 4-NH2 (Scheme 70). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 41

Scheme 69

Scheme 70 42 Ferraboschi, Ciceri, and Grisenti

600 The preparation of the N4-pentyloxycarbonyl derivative seems to be the crucial step of this synthesis. The introduction of the carbamoyl function, according to two Hoffman-La Roche patents, published in 1995,140,141 was postponed to the last step of the synthesis. In one case 141 the N4-functionalization was realized directly on 50-deoxy-5-ﬂuorocytidine 114 605 affording in 92% yield 50-deoxy-20,30-di-O-pentyloxycarbonyl-5-ﬂuoro-N4-pentyloxycar- bonylcytidine 128. The 20,30-carbonates were removed with aqueous sodium hydroxide in methanol (Scheme 71).

Scheme 71

An improvement of the preparation of the N4-pentyloxycarbonyl derivative was obtained by Teva by modifying the reaction conditions:142 choosing 2-methyltetrahydro- 610 furan as solvent, the pyridine amount was signiﬁcantly reduced (1.7 eq), avoiding its removal by evaporation in the course of work-up. After removal of acetates from compound 129 pure capecitabine 63 in 86% yield was obtained (Scheme 72).

Scheme 72

Attempts to introduce the pentyloxycarbonyl group directly on an unprotected 50- deoxy-ﬂuorocytidine led to the development of new reagents, different from the usual 615 pentylchloroformate, able to selectively functionalize the N-4. With this purpose in mind, starting from N-hydroxysuccinimide143, a new pentylcarbonate 130 was recently prepared in 98% yield that selectively functionalized the unprotected 50-deoxy-ﬂuorocytidine at N- 4, affording capecitabine 63 in 44% yield (Scheme 73). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 43

Scheme 73 With the same aim, some pentyloxycarbonylation reagents like hydroxybenzotria- 620 zole-131 and pentaﬂuorophenoxyderivative, 132 were prepared144 and used to selectively form capecitabine 63 from 50-deoxy-ﬂuorocytidine 114, in 66 and 67% yield, respectively (Scheme 74).

Scheme 74

The same authors also published a method145 for the preparation of N-functionalized cytidine that, using pentylchloroformate, required the protection of 20and 30-hydroxy 625 groups, obtained through the formation, by reaction with thionyl chloride, of cyclic sulfinyl ester 133, easily purified by crystallization (85% yield). After the N4-functionalization (85%), the sulfinyl ester of compound 134 was removed by treatment with 1M sodium hydroxide and methanol affording capecitabine 63 (82%) (Scheme 75).

Scheme 75 44 Ferraboschi, Ciceri, and Grisenti

Recently a continuous synthesis146 (Scheme 76) was reported: the synthetic pathway starts 630 from 1,2,3-O-triacetyl-5-deoxy-D-ribose 105 and the glycosylation is catalyzed by a Brønsted acid chosen from among the piridinium triflates. Capecitabine 63 is recovered at the end of the one-flow, multistep synthesis in 72% yield. This method was applied also to other 50-deoxyribo- nucleosides as the previously reported doxifluridine 62 (89%) and galocitabine 135 (see below).

Scheme 76

Since capecitabine 63 is an active pharmaceutical ingredient the identification and 635 the determination of its impurities, as well as the optimization of HPLC methods are very important aspects to be considered together with the synthetic procedures. Some HPLC methods142,147 and some purification processes utilized for the preparation of substan- tially pure capecitabine were reported.148,149 In one of these patents149 is mentioned the possibility of removing the 20,30-O-acetyl groups by means of a hydrolytic enzyme, an 640 unspecified lipase. On the other hand the chemo-enzymatic approach is very attractive for a polyfunctional compound such as capecitabine 63, considering the capability of enzymes to catalyze selective transformations. The enzyme-catalyzed regioselective hydrolytic removal of 50-acetate from nucleoside triacetate 110 (See Scheme 60)112 allowed workers to obtain the 50-hydroxy-20,30di- 645 O-acetyl derivative 111 suitable for the reductive removal of the 50-hydroxy group, necessary to obtain the 50-deoxy nucleoside. In a previous patent150 the same research team reported the results about the selective deprotection of hydroxy groups of nucleosides pol- yesters by means of immobilized lipases; moreover, the topic of regioselective lipase catalyzed transformations of nucleosides is treated exhaustively. In 2008,151 they applied the 650 enzymatic regioselective hydrolysis of 50-O-acetyl group to the 20,30,50-tri-O-acetyl-5-fluorocytidine 136 and to N4-pentyloxycarbonyl-5-fluoro-20,30,50-tri-O-acetyl cytidine 137 (Scheme 77) obtaining advanced intermediates of capecitabine synthesis. High yields were obtained in both 4-NH2 compound 136 (80–97%, immobilized Candida rugosa lipase as enzyme) and N4-derivativative 137. On this last substrate the highest yield 655 (92%) was obtained by using the immobilized protease N from Bacillus subtilis. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 45

Scheme 77

The same regioselective hydrolysis was also studied comparing immobilized recombinant or commercially available lipases from Candida rugosa.152 More recently,153 the chemo-enzymatic approach to a capecitabine intermediate, compound 140, quite similar to that prepared by Cook and co-workers in 1979110 660 (Scheme 78) for doxiﬂuridine 62 synthesis, was studied.

Scheme 78

20,30,50-Tri-O-acetyl-5-ﬂuoro-N4-pentyloxycarbonyl-cytidine 137, substrate of the enzymatic reaction, was prepared as depicted in enzymes, the best results being observed in the case of a cross linked aggregate preparation of the protease from Bacillus licheni- formis (subtilisin), the Alcalase-CLEAÒ: at 96% conversion the 50-acetate was removed 665 almost exclusively (91%). In order to transform advanced intermediate 139 into the ﬁnal capecitabine a careful study about the more suitable 50-functional group and reducing agent was necessary. The 50-deoxynucleoside was obtained starting from 50- 46 Ferraboschi, Ciceri, and Grisenti

Scheme 79

iododerivative 140 by reaction with tributyltin hydride, followed by deprotection of the 20,30-O-diacetyl groups (Scheme 79). The CLEA preparation increases the stability of the 670 enzyme in alcoholic solutions and allows an easier work-up. The work includes the 1H and 13C NMR analyses of the nucleosidic intermediates and of the final capecitabine. Scheme 79, starting from commercially available intermediates. Triacetate 137 was treated, under alcoholysis conditions, with some hydrolytic Finally, another chemo-enzymatic approach was described in 2015.154 In this 675 case, similarly to an already described doxifluridine 62 synthesis112 the biocatalyzed reaction was a transglycosylation reaction of 50-deoxyuridine 113 and the preformed nucleobase of capecitabine, namely N4-pentyloxycarbonyl-5-fluorocytosine 123;the employed enzyme was a pyrimidine nucleoside phosphorylase (Scheme 80). After the introduction in therapy of capecitabine 63, several analogues were synthe- 680 sized and submitted to biological tests with the aim of enhancing its efficacy. Mainly, the modifications were carried out on the N-4 position where different groups, such as for instance BOC155, were introduced instead of the pentylcarbamate. The substitution of an N-4 bonded group can lead to a prodrug that, in hypoxic environments typical of solid tumors, can be reduced by one-electron processes that are 685 inhibited in the normoxic environment of normal tissues. Radiolysis can activate a Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 47

Scheme 80

prodrug to release the active drug after one-electron reduction. An example of a suitable prodrug for radiolysis is N4-nitrothienylderivative 141 described in a 2006 patent (Scheme 81).156

Scheme 81

Some 20,30-O-cyclic carbonates 142 modiﬁed at N-4 showed potent antitumor activ- 690 ity against a leukemia cell line.157 Their synthesis is reported in Scheme 82. 48 Ferraboschi, Ciceri, and Grisenti

Scheme 82 5. Galocitabine Galocitabine (50-deoxy-5-fluoro-N4-(3,4,5-trimethoxybenzoyl)cytidine), (Ro 09–1390) 135, canbeconsideredasacapecitabine derivative with a different N4-substituent or as a doxifluridine prodrug, since at pH 2.1 it rapidly decomposes affording the latter compound.158 695 Galocitabine was synthesized from 20,30-diacetate-50-deoxy-5-fluorocytosine 129 in 1995159 according to Scheme 83. Instead of trimethoxybenzoyl chloride a different derivative of tri- methoxybenzoic acid can be used,144 for example the pentafluorophenoxy ester.

Scheme 83 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 49

The 20,30-O-cyclic sulfinyl ester145 of 50-deoxy-5-fluorocytidine 114 was an alternative protecting group, utilized during the introduction of N4-acyl group. 700 The continuous flow synthesis, like in the case of capecitabine 63, was also applied to the preparation of galocitabine 135 (89%).146

6. 1-Ethoxymethyl-5-Fluorouracil and Emitefur 1-Ethoxymethyl-5-ﬂuorouracyl 143 and emitefur 144 (Scheme 84) have in common the same N1-substitution, namely an ethoxymethyl moiety, which is an acyclic ether, instead 705 of the tetrahydrofuran ring present in tegafur 60. Moreover, emitefur (more commonly called BOF-A2) 144 is also functionalized on N-3. Both were designed to explore 5-ﬂuo- ropyrimidine analogues of nucleosides endowed with improved antitumor activity.

Scheme 84 1-Ethoxymethyl-5-ﬂuorouracyl 143 was synthesized in 1978 in very poor yield (4%) by reaction of ethoxychloromethane with 5-FU 1160 or by reaction of the silylderivative 710 of 5-FU 47 with diethoxymethane and sodium iodide (44% yield)161 or stannic chloride (14% yield),162 as depicted in Scheme 85.

Scheme 85 50 Ferraboschi, Ciceri, and Grisenti

Starting from 1-ethoxymethyl-5-ﬂuorouracil 143 emitefur 144 waspreparedin25% yield (Scheme 86)163,164 by reaction with the isophthaloyl chloride 145 followed by reaction with 6-benzoyloxy-3-cyano-2-pyridone 146. Emitefur 144 is degraded in vivo into 1-ethoxy- 715 methyl-5-ﬂuorouracyl 143, a prodrug of 5-FU, and 3-cyano-2,6-dihydroxypyridine (CNDP) 147, a competitive inhibitor of DPD, the enzyme responsible for rapid degradation of 5-FU 1.5

Scheme 86

7. Carmofur Carmofur 148 is the international non-proprietary name of 1-hexylcarbamoyl-5-ﬂuoro- uracil (HCFU), an antineoplastic agent orally administered in resected colorectal cancer 720 patients. It has some anticancer activity of its own and it is ultimately transformed in vivo to 5-FU 1.165 It was prepared165 starting from 5-FU 1 by reaction with hexyl isocyanate in pyridine or in dimethylacetamide solution, or by reaction with phosgene followed by treatment with 1-hexylamine (Scheme 87). The hexyl isocyanate was, in turn, prepared starting from heptanoyl chloride by reaction with sodium azide,165 or by treatment of 1- 725 hexylamine with trichloromethyl chloroformate, followed by distillation.166 In a polar solvent, like dimethylsulfoxide, carmofur 148 exists as two mixed structures (presumably a hydrogen-bonded structure and a non-hydrogen-bonded structure) at room 1 166 temperature, as demonstrated by H NMR analyses in DMSO-d6 at room temperature and at 80 C or in CDCl3. Deuterated derivatives of carmofur bearing deuterium atoms on the 730 hexyl chain were prepared in order to evaluate the isotopic effect on the metabolism rate.165 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 51

Scheme 87

8. Related Non-Clinical Fluoropyrimidine Nucleosides New syntheses of fluoropyrimidines nucleosides are usually aimed at the preparation of new potential antitumor, antiviral or antibacterial compounds and their destiny depends on the results of clinical investigations. The fluorination in most cases takes place at the 735 5-position, namely the position involved in the thymidine biosynthesis. In the previously examined compounds the fluorine atom is directly linked to C-5 of heteroaromatic ring. In the case of 5-trifluorothymidine (TFT) 149 a trifluoromethyl group instead of the 5- methyl group is present.

Scheme 88 52 Ferraboschi, Ciceri, and Grisenti

TFT 149 was synthesized in 1964 by Heidelberger167 according to Scheme 88, but its 740 development as antineoplastic agent was early hampered by its short half-life (12 min).168 The renewed recent interest for TFT 149, prompted the synthesis of a convenient prodrug. In a 2016 patent 169 the silylation of 30 and 50-hydroxy group of TFT was realized in order to prolong the TFT 149 circulatory half-life. Two methods of silylation affording compound 150 (Scheme 89) were described.

Scheme 89

745 III. Pyrimidine Nucleosides Fluorinated at the Sugar Moiety Fluorine atom introduction at nucleosides can occur not only at the pyrimidine moiety but also at the glycone moiety, modifying the biological activity to an equally significant extent, leading to antiviral or antitumoral molecules. The sugar can be fluorinated at 20,30,4,50-positions or can bear an exocyclic fluorocarbon substituent. Considering that 750 some reviews170–173 already cover the topic of nucleosides fluorinated at the sugar moiety, we focused on an antitumor pyrimidine nucleosides difluorinated at the 20-position, namely gemcitabine. The presence of fluorine at the 20-position can provide acidic and enzymatic stabilities to the glycosidic bond174 and the gem-difluoromethylene group is an isopolar and isosteric substituent for oxygen.175 Extensive studies about the 20-position 755 modification with atoms or groups other than hydrogen or hydroxy (for example Cl,176, 177 178 179,180 N3, Se, CH3, ) have been performed, since the presence of a hydroxy or a hydrogen distinguishes nucleosides as components of RNA or DNA.

1. Gemcitabine Gemcitabine 151 hydrochloride (20,20-diﬂuoro-20-deoxycytidine hydrochloride) is cur- 760 rently produced and marketed as GemzarÒ for the treatment of various cancers such as pancreatic, breast, non-small cell lung and ovarian cancers. It is administered by intravenous infusion, the oral use of gemcitabine being limited by poor bioavailability. Gemcitabine is intracellularly converted by deoxycytidine kinase into its therapeutically active metabolites, the 50-diphosphate 152 and the 50-triphosphate derivatives 153 765 (Scheme 90). 50-Diphosphate 152 inhibits the ribonucleotide nuclease, the enzyme responsible for regulating the total rate of DNA synthesis, while 50-triphosphate 153 can become incorporated into the DNA and inhibits the nuclear replication. The 50-position is directly involved in the action of gemcitabine 151, while the 4-NH2 group plays a role in the Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 53

Scheme 90

770 enzyme-substrate recognition through the formation of hydrogen bonding interactions with the kinase; the removal of the amino group by the liver enzyme cytidine deami- nase117 leads to the inactive metabolite 20,20-difluorouridine. The preparation of the 20,20-difluorosugar is the main target in gemcitabine 151 synthesis and many different approaches have been proposed (some of that described in a 775 minireview181) since the first synthesis reported by Hertel in 1988.182 The starting material for this synthesis (Scheme 91) was protected (R)-glyceraldehyde 154 prepared from D-mannitol; its reaction with ethyl bromodifluoroacetate led to a diastereomeric mixture of (3R)- and (3S)-propionate (3R/3S 3:1) derivatives 155 that were separated by silica gel column chromatography. The (3R)-isomer, 780 under acidic conditions afforded lactone 156 that, after protection of 3- and 5- hydroxy groups (compound 157), was reduced with diisobutylaluminum hydride to a mixture of a-andb-anomers (158); the anomeric mixture was transformed into the corresponding 1-O-mesylate (compound 159) necessary for the substitution by the silyl derivative of cytidine. The formation of the N-glycosidic bond was the crucial 785 step of the synthesis since the required b-anomer of 151 was the minor product (10%). The undesired a-anomer was obtained in 40% yield, after removal of the protecting group and separation by reverse phase column chromatography. An improvement to this synthesis was achieved by Chou and co-workers,183 some years later, through the selective crystallization of the diastereomeric mixture 790 of 3,5-di-O-benzoyl-difluororibono-lactone 161 obtained by esterification of 30-benzoate 160 (Scheme 92). The choice of benzoyl as protecting group of 3- and 5- hydroxy groups allowed selective crystallization of the desired erythro-isomer. Reduction of the 1-carbonyl group afforded a mixture of a-andb-anomers 161 that was transformed into the corresponding mesylates 162. The glycosylation reac- 795 tion afforded a 1:1 a/b anomeric mixture of nucleosides 163 instead of the 4:1 mixture observed by Hertel which used tert-butyldimethylsilyl ether as protecting group. Starting from protected (R)-glyceraldehyde 154 many modifications, aimed to improve the yield and the purity of the final b-anomer, were proposed relating to the 54 Ferraboschi, Ciceri, and Grisenti

Scheme 91

800 use of different reducing agents of the lactone, to the introduction of alternative protecting groups, to the transformation of 1-hydroxy into a different leaving group and, finally, to the purification process of the obtained gemcitabine 151.Morethan ten patents have been published in 2005–2015 about this topic. For example, an Indian team184,185 was able to obtain 3,5-di-O-benzoyl-2,2-difluoro-lactone with a 805 high purity (99.8% erythro-isomer) by treatment of the diastereomeric mixtures of lactones 164, containing also the corresponding hydroxyacid 165, with para-toluen- sulfonic acid in toluene, at reflux, removing the water by means of the Dean-Stark apparatus, followed by crystallization (Scheme 93). Reduction with VitrideÒ (sodium bis(2-methoxy)aluminum hydride, also known as Red-AlÒ or SMEAH) followed by 810 treatment with mesyl chloride afforded usual intermediate 162 for the accomplish- ment of the synthesis. The gemcitabine hydrochloride showed a 95% diastereomeric purity that was increased to 99.9% by means of crystallization by water/acetone or water/2-propanol. Another approach186 made use of DIBAL as reducing agent of protected lactone 166 815 and mesylate as leaving group in compound 167, postponing all purification processes until the end of the synthesis. By this method, 92.6% pure gemcitabine was obtained after Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 55

Scheme 92

column chromatography. The final crystallization of gemcitabine hydrochloride from water/acetone afforded 99.6% pure gemcitabine hydrochloride (Scheme 94). The choice of a suitable protecting group187 for the 3- and 5-hydroxys (benzoyl 820 substituted with an electron withdrawing group), with the aim to obtain solid compounds, allowed preparation of the pure erythro-lactone 168 by crystallization from ethyl acetate/ hexane. The synthesis was then finished in the usual way; the precipitation of gemcitabine hydrochloride afforded 99.9% pure product (Scheme 95). Tri-tert-butoxyaluminum hydride was used in the course of the reduction of the lactone 825 protected as 3,5-di-O-(3-fluorobenzoyl) ester187 or 3,5-di-O-(4-phenyl)benzoyl ester.188 By means of Red-AlÒ was instead obtained the lactol, protected at the hydroxy groups as tert- butyldimethylsilyl ethers189 or benzoates,190,191 or naphthoyl esters,192 or as 3-O-benzoate and 5-carbamate.193 In this last case, the trichloroacetimidate was chosen as the leaving group. If the same leaving group was introduced and the obtained trichloracetimidate directly 830 submitted to the reaction with protected cytidine, in a “one-pot” process,190 the protected 56 Ferraboschi, Ciceri, and Grisenti

Scheme 93

Scheme 94 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 57

Scheme 95 nucleoside was obtained in high yield (89%); gemcitabine hydrochloride was obtained (99.94% b-anomer) in 30% yield after deprotection and puriﬁcation process. The use of the cinnamoyl ester as protecting group of 3- and 5- hydroxyl allowed the selective precipitation of an anomer at different steps of the synthesis.194,195 835 In order to avoid the energy waste required to maintain the low temperatures (between ¡80 and ¡60C) during the reduction of the 1-carbonyl group with DIBAL, in a 2010 patent the use of calcium borohydride in tetrahydrofuran (90% yield) or sodium borohydride in ethyl acetate (88% yield) was proposed.196 In order to increase the b/a anomer ratio, 1-iodido-2,2-diﬂuoro derivative 169, 840 as substrate of the N-glycosylation, was prepared through two alternative ways: directly from 1-hydroxy 170 by reaction with iodine and triphenylphosphine in dichloromethane, or through its mesylate, with sodium iodide in acetone. The addition of a silver salt to the glycosylation reaction mixture led to the formation of oxo- nium intermediate, 171 that underwent the attack of nucleobase from the top of the 845 molecule, favoring the formation of the b-anomer (a/b 1:5.6).197 Inamorerecent patent198 the authors reported that even more favorable ratios can be obtained adding 58 Ferraboschi, Ciceri, and Grisenti

an oxidizer; for example, a 1:18 a/b anomer ratio was observed using (NH4)2S2O8 in acetonitrile. (Scheme 96).

Scheme 96

The presence of a bulky protecting group (para-phenylbenzoate)199,200 of the 3- 850 hydroxy group allowed workers to obtain a solid 1-phosphate, compound 172 showing a 1:10.8 a/b ratio, that by a simple crystallization afforded in 77% yield the b-anomer (98.24% pure). The b-phosphate was easily converted into the corresponding a-bromo derivative 173, which is a suitable precursor, after crystallization, of the N-glycosylated product with the desired b-stereochemistry. In order to avoid pos- 855 sible anomerization of a-bromo sugar 173 the trimethylsilyl bromide, generated in Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 59

the course of substitution with the protected nucleobase, was removed by distillation using heptane as carrier. By this method, glycosylation occurred in high yields (92%) giving compound 174 showing a 1:5.5 a/b ratio. Pure gemcitabine 151 was obtained, after removal of protecting groups and crystallization in 65% yield from 860 a-bromo sugar 173 (Scheme 97).

Scheme 97 60 Ferraboschi, Ciceri, and Grisenti

According to a different approach, instead of the usual protected (R)-glyceraldehyde 154, 1,6-anhydro-b-D-glucose 175, prepared by thermal degradation of starch, was used as starting material.201 2,4-Trimethylsilyl ether 176 wasoxidizedto3-ketoderivative177 by means of Dess-Mar- 865 tin reagent. After removal of the silyl ether the 2- and 4-hydroxy groups were transformed into corresponding methyl ethers 178 (96% yield) and the 3-ketoderivative was converted into gem- difluoro derivative 179 by reaction with diethylaminosulfuryl trifluoride (DAST) and N,N- dimethylpropylene urea-hydrogen fluoride (DMPU-HF) (90% yield). Opening the anhydro ring under acidic conditions (compound 180) and oxidation with sodium periodate afforded 2- 870 deoxy-2,2-difluoro-D-ribose 181 in 68% yield. The synthesis was accomplished as reported in Scheme 98. 1,3,5-Tri-O-acetyl-D-ribose difluoroderivative 182 was condensed with the N1-cis- (2-cyanovinyl)-N,N-bistrimethylsilyl urea and the obtained intermediate was cyclized under strong alkaline conditions to the final b-nucleoside in 11% yield, after column chromatography. The overall yield from 1,6-anhydro-b-D-glucose 175 was 4.4%, confirming the difficulties 875 encountered in other synthetic approaches, deriving from the presence of two stereogenic centers, in the 1- and 3-positions, generated or modified in the course of the synthesis.

Scheme 98

The direct ﬂuorination of a preformed nucleoside can overcome the formation of the erythro and threo-diastereoisomers encountered starting from (R)-glyceraldehyde 154;in addition, the N-glycosylation with a non-ﬂuorinated sugar can afford a more favorable 880 a/b anomers ratio. To this end, a suitably protected 20-ketonucleoside 183 was prepared starting from the commercially available cytidine and submitted to the reaction with Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 61

DAST in presence of pyridine-hydrogen fluoride (Scheme 99).202 N4,30,50-Tri-(ortho- toluyl)-20-ketocytidine 183 was prepared in 53% overall yield from cytidine. Reaction with the fluorinating agent afforded protected difluoronucleoside 184 (80% from HPLC 885 and NMR) that was converted into gemcitabine 151 by removal of protecting groups.

Scheme 99

In another example 2-deoxy-D-ribonolactone 185 was chosen as the substrate of the fluorination, which was carried out in two steps with N-fluorodibenzenesulfonimide (NFSi) in presence of lithium bis(trimethylsilyl)amide (LiHMDS) (Scheme 100).203 Pro- tected gem-difluororibonolactone 186, obtained in 47% yield from 2-deoxy-D-ribonolac- 890 tone 185, was then transformed into gemcitabine 151 as depicted in Scheme 100. In 2014, a team of Merck researchers published the synthesis of 2-deoxy-2,2-difluoro-D- ribose according to their “de novo” approach (Scheme 101).204 Optically active aldehyde 187 (83% ee) prepared through an enantioselective method, developed by the authors, was coupled with isopropyl bromodifluoroacetate under Reformatsky conditions: the suitably functionalized 895 pentanoate was obtained as a mixture of diastereoisomers 188.Purificationbyflashchromatog- raphy provided the required diastereoisomer in 59% yield. Removal of the 2,2,6,6-tetramethyl- piperidinyl group (TMP) and concomitant cyclization afforded the difluorolactone 189 precursor of the ribose ring. Gemcitabine synthesis was then carried out through 1-iodine derivative 190, in the presence of persulfate, as previously reported.197,198 (see Scheme 96). At the end 900 of the synthetic sequence gemcitabine was isolated in 23% overall yield, from optically active aldehyde 187,anda1:4a/b-anomers ratio. This synthesis offers a promising approach to gemcitabine preparation; indeed the final purification could be carried out according to published methods. 62 Ferraboschi, Ciceri, and Grisenti

Scheme 100

Beside the synthetic work, there are some publications dedicated exclusively to the 905 puriﬁcation of gemcitabine base or hydrochloride. For example 99.8% pure gemcitabine hydrochloride was obtained by fractional crystallization at acidic pH ( 0.5).205

2. Gemcitabine Derivatives Gemcitabine chemotherapeutic efﬁcacy is limited by its high toxicity to normal cells and short plasma half-life (9–13 min for human plasma) dependingontherapiddeaminationbycytidine 910 deaminase in the liver, kidneys and plasma to less cytotoxic metabolites. Various prodrug strategies have been developed to overcome these adverse aspects and to allow for oral delivery. Protection of the 4-amino group, for example as the amide, can facilitate a slower release of gemcitabine, increasing the bioavailability and uptake and providing resistance to enzymatic deamination. A series of 4-N-alkyl and 4-N-alkanoyl compounds were pre- 915 pared in 2014206 and their activity evaluated. N-alkanoyl gemcitabine derivatives 191 showed potent cytostatic activity in the nanomolar range, whereas N-alkyl derivatives 192 required micromolar range. The N-alkanoyl derivatives 191 were prepared in 40– 66% yield by reacting gemcitabine with a carboxylic acid in dimethylsulfoxide, in presence of N-methylmorpholine, 1-hydroxybenzotriazole and (N-dimethylaminopropyl)-N1- 920 ethyl-carbodiimide (Scheme 102). The N4-alkyl derivatives 192 were prepared from protected gemcitabine 193 through its N4-tosylate by reaction with an amine. N4-(2-Propyl-1-oxopentyl) derivative 194 was prepared in 2006,207 by Eli Lilly (Scheme 103), designed to be resistant to deamination by hydrolysis under acidic conditions similar to those found in the digestive system but able to release gemcitabine 925 upon action by carboxylesterase. Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 63

Scheme 101

N4-Benzoyl derivative 195 was prepared as depicted as shown in Scheme 104 and included into oligonucleotides aimed to preferentially kill cancerous cells over non-cancerous cells.208 Some oligonucleotides containing gemcitabine derivatives were shown to be more effective in killing cancerous cells at equivalent dosages of gemcitabine itself. 930 As in the case of capecitabine156 the hypoxic conditions of tumor cells can selectively activate a prodrug, mediating the fragmentation of a masked cytotoxic compound into the active cytotoxic agent; the masked cytotoxic agent can be N4-carbamate 196156 (Scheme 105)ora50-ester with lipoic acid derivative 197209 (Scheme 106) that plays the role of the redox-modulating agent. 935 In a 2013 article the conjugation of gemcitabine with coumarin derivative 198 and biotin derivative 199 was described:210 the molecule contains the cleavable disulﬁde group of biotin, a molecule taken up preferentially by cancer cells, and a ﬂuorescent moiety to enable real-time monitoring of the drug delivery. The synthesis of conjugate 200 and the proposed mechanism of action under physiological conditions, involving GSH, 940 the most abundant thiol in the cells, are depicted in Scheme 107. 64 Ferraboschi, Ciceri, and Grisenti

Scheme 102

Scheme 103 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 65

Scheme 104

Besides conjugation with polymeric materials,211,212 the preparation of N4-derivatives with lipoaminoacid residues, suitable for penetration through biological membranes and barriers, was also reported.213 Lipoaminoacids are aminoacids bearing an alkyl chain in the 2-position. They impart to the molecule to which they are conjugated amphiphilic 945 properties. Lipoaminoacid derivatives 201, prepared in 69% yield from gemcitabine hydrochloride (Scheme 108), in vitro results were encouraging of the study of the loading of these derivatives in liposomes and in other lipid based drug carriers. Controlled drug delivery from chiral molecules is the aim of a study about the possibility of incorporating gemcitabine into the enantiomers of bis(diamido)-bridged basket 950 resorcin[4]arene 202,214 in the gaseous phase (Scheme 109). The diastereomeric complexes among the macrocycle, gemcitabine and a chiral amine behave as supramolecular devices which, depending on the configuration of macrocycle and amine, can or cannot release the nucleoside. Complexation phenomena were investigated215 by NMR methods and molecular modeling, allowing the identification of two different interaction sites of 955 the guest in the resorcin[4]arene host. Different behaviors of resorcin[4]arene complexes with 20-deoxycytidine, its 20-epimer cytarabine and gemcitabine were observed. Accord- ing to the authors, the assessment of the factors regulating the formations of these complexes offers the possibility to modulate the drug/receptor interactions, through the electronic properties of 20-substituents on the nucleoside furanose ring. 960 The bioorthogonal approach, already described in the case of floxuridine,105 was also applied by the same authors to the preparation of gemcitabine prodrugs.216 Between the two masking possibilities, 50-O-carbonate or N4-carbamate, the latter was chosen due to its higher 66 Ferraboschi, Ciceri, and Grisenti

Scheme 105

stability to enzymatic metabolism. Starting from gemcitabine 151, propargyloxycarbonyl derivative 203 was prepared (Scheme 110), and treated with a Pd0-functionalized resin under 965 biocompatible conditions (buffer pH 7.4, 37C), in order to remove the carbamate, and, then, placed in cell culture. In both cases the formation of free gemcitabine was monitored. DNA damage observed in pancreatic cancer cell cultures treated with N4-functionalized gemcitabine and Pd0-resin conﬁrmed the bioorthogonal generation of free gemcitabine 151.

3. Related Non-Clinical Pyrimidine Nucleosides Fluorinated at the Sugar Moiety 970 20,20-Difluorinated azacytidine 204 and 20,20-difluorodeoxyribose-trifluorothymidine 205 were recently169 prepared starting from suitably protected 1-bromo-20,20-difluorosugar, compound 206, in turn prepared as depicted in Scheme 111 and silylated (see Scheme 88) in order to obtain new fluoronucleosides. Compound 204 is noteworthy for the presence of a nitrogen atom at the 5-position, 975 instead of a carbon, in addition to other modifications. Future studies will establish the clinical usefulness of these modifications. In 2015217 two 20-fluorotricyclo-DNA nucleosides 207 and 208 were synthesized. The introduction of fluorine on methylglycoside 209 was achieved through the electrophilic addition of Selectfluor to the 1,2-double bond of 210 and an approximately equimo- 980 lar mixture of 2-ribo 211 and 2-arabino 212 fluoro sugars was obtained as established by 1H NMR NOE studies (Scheme 112). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 67

Scheme 106

The highly selective N-glycosilation of 2-ribosugar 211, obtained after column chromatography, afforded thymine 20-ﬂuoro nucleoside 207 showing a 1:12 a/b anomeric ratio (Scheme 113). 985 The same intermediate 213 was converted into nucleoside 208 bearing the 5-methyl- citosine as nucleobase according to Scheme 114. Compounds 207 and 208 represent two novel ﬂuorinated nucleoside analogues which in the future will be incorporated into oligonucleotide backbone structures.

IV. Radiolabeled Fluoropyrimidine Nucleosides as PET Tracers 990 Radiolabeled antitumor agents may provide useful information about their metabolism and pharmacokinetics, in vivo, using functional imaging modalities such as Positron Emission Tomography (PET). Indeed, due to its good spatial resolution, PET allows workers to map the distribution of the radiotracers and their biological targets; moreover, depending on the selected molecular structure, PET is a powerful tool to accurately measure metabolic factors 995 and/or biochemical and physiological parameters. In the last decades, PET has found its major applications in oncology, where the preparation of molecules labeled with positron emitting radionuclides, and their subsequent use with human subjects provides useful diag- nostic information. It even allows clinicians to directly affect therapeutic strategies. Based on the detection of high energy gamma photons generated by the annihilation of the posi- 1000 tron-electron pairs, PET is a highly sensitive technique, and radiolabeled tracers in nano to 68 Ferraboschi, Ciceri, and Grisenti

Scheme 107 picomolar concentrations (non-pharmacological concentrations) may be used, thus minimiz- 11 ing their potential toxicity. The most useful PET radionuclides are the short half-life C(t1/2 18 11 D 20.4 min) and F(t1/2 D 109.6 min). The ﬁrst nucleoside based PET tracer was [ C]- Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 69

Scheme 108

thymidine218,andlater[11C]-5-FU was synthesized219, with the aim to monitor the levels of 1005 5-FU following therapeutic treatments. Starting from the b-(N-Benzoylamino)-a-ﬂuoroacry- lamide 214, prepared from sodium ethyl-2-ﬂuoro-3-hydroxyacrylate 215,asdepictedin Scheme 115, the introduction of 11C was achieved using [11C]-phosgene, prepared by the same authors in 2002.220 Acrylamide with the suitable double bond E-stereochemistry 216 was obtained by the irradiation of the Z-isomer 217. The radiochemical yield was about 25%. 1010 11C was initially selected as the preferred radionuclide because it replace a 12C in the molecular structure, thus not altering the biological behavior of the radiolabeled probe, compared with its “cold” counterpart. On the contrary, its very short half-life may be a

Scheme 109 70 Ferraboschi, Ciceri, and Grisenti

Scheme 110

serious drawback, especially in case of slow kinetics, in vivo. For these reasons, attention was then focused on the preparation of F-18 labeled radiotracers. F-18 has a considerably 1015 longer half-life, which is compatible with the biological half-life of slow kinetics molecules, and at the same time allows for longer radiosynthetic pathways. Furthermore, F-18 is produced efficiently and with high yield, and higher starting activity may also help to improve the radiopharmaceutical’s availability, Finally, the 110 min half-life of F-18 allows for the radiotracer distribution to other PET centers not equipped with a cyclotron, 1020 thus improving availability and reducing general preparation costs. For instance, the [11C]-based tracers were later superseded by [18F]-FLT (30-deoxy-30-fluorothymidine),221 which is still one of the most frequently used PET radiotracers in the field of oncology. [18F]-capecitabine218222, 223 was synthesized, as an enzyme based imaging agent, to enable non- invasive monitoring of tumor enzymes (thymidine phosphorylase and uridine phos- 1025 phorylase) (see Scheme 1) and tumor response to capecitabine therapy, using PET technique. A 2004222 paper reported the synthesis of [18F]-capecitabine 218 through the nucleophilic substitution of the 5-nitro group of 20,30-di-O-acetyl-50-deoxy-5-nitro-N4-(pentyloxycarbonyl)cytidine 219 by reaction with K18F (prepared in a RDS-112 cyclotron) in the presence of a bicyclic kryptand, Kryptofix 2.2.2 (Scheme 116). The nitro-precursor 1030 was prepared from 5-deoxy-1,2,3-tri-O-acetyl-D-ribose 105 (obtained from D-ribose) and 5-nitrocytosine 220 in 18% overall yield. A different approach was described by a Korean team223 using as substrate N4-(pentyloxycarbonyl)-cytidine 221, which introduced 18F by direct electrophilic fluorination 18 with [ F]-F2 generated in a MC-50 cyclotron. The radiochemical yield and radiochemi- 1035 cal purity were 5–15% and >95%, respectively (Scheme 117). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 71

Scheme 111

An attempt to prepare a [18F]-labeled gemcitabine derivative,206 failed at the early steps of the synthesis since the addition of unlabeled HF (70% in pyridine) to N4-(10- undecenoyl)-20,20-difluorocytidine led to a mixture of 10-, 9- and 8-fluorinated isomers. A very interesting enzyme-catalyzed formation of C-F, suitable also to produce 1040 labeled fluoronucleosides was proposed by O’Hagan and coworker in 2003.224–228 They used the enzyme fluorinase, crude or partially purified, extracted from Streptomyces cat- tleya. The enzyme is capable to catalyze the synthesis of fluoro compounds, particularly fluoronucleosides, in the presence of a fluorine donor (lithium, potassium or sodium fluoride) or hydrogen fluoride and of the substrate. (Scheme 118). 1045 The substrate was S-adenosyl-L-methionine (SAM) 222 that, by action of fluorinase, underwent a nucleophilic substitution (L-methionine 223 as leaving group) leading to 50- 72 Ferraboschi, Ciceri, and Grisenti

Scheme 112

Scheme 113 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 73

Scheme 114

Scheme 115

Scheme 116 74 Ferraboschi, Ciceri, and Grisenti

Scheme 117

Scheme 118 Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 75

fluoro-50-deoxyadenosine 224; the method was also applied to the synthesis of [18F]-50- fluoro-50-deoxyadenosine.227 Hydrolysis of the N-glycosidic bond by means of nucleoside hydrolase from Trypanosoma vivax afforded 5-deoxy-5-[18F]-ribose.221 Starting 1050 from [18F]-50-fluoro-50-deoxyadenosine by means of an enzyme-catalyzed transglycosylation [18F]- 50-fluoro-50-deoxy-5-fluorouracil was also obtained.227 This and other exam- ples recently reviewed229 are promising of a wider application of the method, at the present limited to the fluorination of the 50-position. More recently an enzyme with the same activity was isolated from the marine derived bacterium Streptomyces 1055 xinghaiensis.230

Conclusion Fluorinated nucleosides represent an important category of nucleoside analogues endowed with antitumor and/or antiviral activity. This review focused on antitumor pyrimidine nucleosides bearing fluorine on the nucleobase or on the glycone. 1060 5-FU 1 is more than 50 years old but it is still used, in combination with other anticancer drugs (for example, avastatatin and oxaliplatin), for the treatment of metastatic colorectal cancer. A more recent therapeutic indication of 5-FU 1 is the topical use (cream or solution) for the treatment of skin cancers and Bowen’s disease. Capecitabine (Xeloda) 63 is the most frequently used therapeutic fluoropyrimidine nucleoside, alone or in com- 1065 bination with other drugs and, like other relevant fluoropyrimidines nucleosides analogues, can be regarded as a 5-FU prodrug. Gemcitabine (Gemzar) 151 belongs to the group of nucleosides fluorinated on the glycone moiety and it is among the ten leading brands in the global cancer market. Considering the impressive number of fluorinated pyrimidine nucleosides, a choice 1070 of scope was mandatory, and we decided to select clinically well-established compounds, and their derivatives designed to improve the pharmacological properties of the parent compounds. This choice excluded, for example, nucleoside analogues characterized by the presence of a heteroatom (N, S, Se), different from oxygen in the 5-atoms ring173 or the carbocyclic analogues,172,231 which are no less interesting than the discussed com- 1075 pounds. Despite the numerous compounds that have received marketing authorization, the demand for antitumor oral fluoro pyrimidine nucleosides is still of concern; the design and the synthesis of more selective molecules remains a target of pharmaceutical research. Such research may develop the possibility of patient-tailored therapy based on individualized combinations of the described molecules. We hope that the synthetic meth- 1080 ods described in this review, suitably adapted, could be a valuable aid in the development of new nucleoside analogues.

Acknowledgments We thank Dr. Paola Rota of the Department of Scienze Biomediche, Chirurgiche e Odon- toiatriche for technical assistance.

1085 References 1. T. Ishikawa, M. Utoh, N. Sawada, M. Nishida, Y. Fukase, F. Sekiguchi and H. Ishitsuka, Bio- chem. Pharmacol., 55, 1091 (1998). 76 Ferraboschi, Ciceri, and Grisenti

2. A. Yokoyama, Y. Nakai, S. Yoneda, Y. Kurita and H. Niitani, Anti-cancer Drugs, 8, 574 (1997). 1090 3. R. J. Rutman, A. Cantarow and K. E. Paschkis, Cancer Res., 14, 119 (1954). 4. C. Heidelberger, N. K. Chaudhuri, P. Danneberg, D. Mooren, L. Griesbach, R. Duschinsky, R. J. Schnitzer, E. Pleven and J. Scheiner, Nature, 179, 663 (1957). 5. M. Malet-Martino, P. Jolimaitre and R. Martino, Curr. Med. Chem. Anti-cancer Agents, 2, 267 (2002). 1095 6. M. G. Rose, M. P. Farrell and J. C. Schmitz, Clin. Colorectal Cancer, 1, 220 (2002). 7. D. Papamichael, Stem. Cells, 18, 166 (2000). 8. C. Aschele, A. Sobrero, M. A. Faderan and J. R. Bertino, Cancer Res., 52, 1855 (1992). 9. A. F. Sobrero, C. Aschele, A. P. Guglielmi, A. M. Mori, G. G. Melioli, R. Rosso and J. R. Ber- tino, J. Natl. Cancer Inst., 85, 1937 (1993). 1100 10. A. F. Sobrero, C. Aschele and J. R. Bertino, J. Clin. Oncol., 15, 368 (1997). 11. R. Duschinsky, E. Pleven and C. Heidelberger, J. Am. Chem. Soc., 79, 4559 (1957). 12. C. Heidelberger and R. Duschinsky, US 2802005, 1957; Chem. Abstr., 52, 11450 (1958). 13. M. J. Robins and S. R. Naik, J. Am. Chem. Soc., 93, 5277 (1971). 14. D. H. Barton, R. H. Hesse, H. T. Toh and M. M. Pechet, J. Org. Chem., 37, 329 (1972). 1105 15. M. J. Robins, M. MacCoss, S. R. Naik and G. Ramani, J. Am. Chem. Soc., 98, 7381 (1976). 16. D. Cech, H. Meinert, G. Etzold and P. Langen, J. Prakt. Chem., 315, 149 (1973). 17. R. Anderson and P. D. Schuman, US 3954759, 1976; Chem. Abstr., 85, 46744 (1976). 18. S. S. Yemul, H. B. Kagan and R. Setton, Tetrahedron Lett., 21, 277 (1980). 19. S. Misaki, S. Ishii, N. Suzuki, M. Wakabayashi and T. Sowa, US 4122251, 1978; Chem. 1110 Abstr., 88, 121236 (1978). 20. A. Yamazaki, H. Morisawa, Y. Oda and K. Uchida, US 4186266, 1980; Chem. Abstr., 90, 121648 (1979). 21. O. Miyashita, K. Matsumura, H. Shimadzu and N. Hashimoto, Chem. Pharm. Bull., 29, 3181 (1981). 1115 22. O. Miyashita, K. Matsumura, H. Shimazu and N. Hashimoto, US 4329460, 1982; Chem. Abstr., 88, 105397 (1978). 23. R. J. De Pasquale and P. D. Schuman, US 4299961, 1981; Chem. Abstr., 95, 187297 (1981). 24. B. Baasner and E. Klauke, J. Fluorine Chem., 45, 417 (1989). 25. R. E. Banks, D. S. Field and R. N. Haszeldine, J. Chem. Soc., C, 1280 (1970). 1120 26. A. Haas and D. Kortmann, Chem. Ber., 114, 1176 (1981). 27. T. Umemoto and E. Ogura, US 4631342, 1986; Chem. Abstr., 104, 34082 (1986). 28. M. Diksic, S. Farrokhzad and L. D. Colebrook, Can. J. Chem., 64, 424 (1986). 29. G. W. M. Visser, S. Boele, B. W. Van Halteren, G. H. J. N. Knops, J. D. M. Herscheid, G. A. Brinkman and A. Hoekstra, J. Org. Chem., 51, 1466 (1986). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 77

1125 30. G. W. M. Visser, R. E. Herder, F. J. J. Dekanter and J. D. M. Herscheid, J. Chem. Soc. Perkin Trans. I, 1203 (1988). 31. G. W. M. Visser, R. Wedzinga, R. P. Klok and J. D. M. Herscheid, J. Chem. Soc. Perkin Trans. II, 231 (1994). 32. G. S. Lal, W. Pastore and R. Pesaresi, J. Org. Chem., 60, 7340 (1995). 1130 33. W. K. Chung, J. H. Chung and K. A. Watanabe, J. Heterocycl. Chem., 20, 457 (1983). 34. T. Fuchikami, A. Yamanouchi and Y. Suzuki, Chem. Lett., 1573 (1984). 35. G. Q. Shi, Q. Wang and M. Schlosser, Tetrahedron, 52, 4403 (1996). 36. D. S. Hitchcock, A. A. Fedorov, E. V. Fedorov, S. C. Almo and F. M. Raushel, Biochemistry, 53, 7426 (2014). 1135 37. M. Kawase, K. Samejima, M. Okada, K. Ochi and I. Matsunaga, Chem. Pharm. Bull., 33, 2395 (1985). 38. F. M. Menger and M. J. Rourk, J. Org. Chem., 62, 9083 (1997). 39. T. W. B. Cai, X. P. Tang, J. Nagorski, P. G. Brauschweiger and P. G. Wang, Bioorgan. Med. Chem., 11, 4971 (2003). 1140 40. J. E. Saavedra, T. R. Billiar, D. L. Williams, Y. M. Kim, S. C. Watkins and L. K. Keefer, J. Med. Chem., 40, 1947 (1997). 41. T. Kametani, K. Kigasawa, M. Hiiragi, K. Wakisaka, S. Haga, Y. Nagamatsu, H. Sugi, K. Fukawa, O. Irino, T. Yamamoto, N. Nishimura, A. Taguchi, T. Okada and M. Nakayama, J. Med. Chem., 23, 1324 (1980). 1145 42. A. Buur and H. Bundgaard, J. Pharm. Sci., 75, 522 (1986). 43. R. M. Phelan, M. Ostermeier and C. A. Townsend, Bioorg. Med. Chem. Lett., 19, 1261 (2009). 44. T. Kametani, K. Kigasawa, M. Hiiragi, K. Wakisaka, K. Nakazato, K. Ichikawa, K. Fukawa, O. Irino, N. Nishimura and T. Okada, J. Med. Chem., 25, 1219 (1982). 45. R. A. Zhuk, A. S. Ludzisha, Y. G. Shpaer, M. Y. Lidak, A. A. Zidermane and D. V. Meirena, 1150 Chem. Heterocycl. Compd., 21, 1159 (1985). 46. Z. S. Zhang, Q. B. Zhang, J. Wang, X. L. Shi, J. J. Zhang and H. F. Song, Carbohyd. Polym., 79, 628 (2010). 47. M. K. Kilevitsa, Yu. A. Maurin’sh, R. A. Paegle, E. Liepin’sh, A. A. Zidermane and M. Yu. Lidak, Khim. Geterotsikl. Soedin., 11, 1532 (1981); Chem. Abstr., 96, 85922 (1982). 1155 48. Yu. A. Maurin’ sh, R. A. Paegle, K. V. Khagi, D. V. Meirena, B. Z. Simkhovich and M. Yu. Lidak, Pharm. Chem. J., 22, 466 (1988). 49. T. Kuroda, K. Hisamura, I. Matsukuma, H. Nishikawa and N. Nakamizo, Bull. Chem. Soc. Jpn., 62, 674 (1989). 50. S. Nishimoto, H. Hatta, H. Ueshima and T. Kagiya, J. Med. Chem., 35, 2711 (1992). 1160 51. M. Mori, H. Hatta and S. Nishimoto, J. Org. Chem., 65, 4641 (2000). 52. S. S. Agasti, A. Chompoosor, C. C. You, P. Ghosh, C. K. Kim and V. M. Rotello, J. Am. Chem. Soc., 131, 5728 (2009). 53. K. Fujimoto, Y. K. Takematsu, A. Shigeno, M. Furusawa and T. Sakamoto, Bioorg. Med. Chem. Lett., 24, 3736 (2014). 78 Ferraboschi, Ciceri, and Grisenti

1165 54. J. S. de Bono and C. J. Twelves, Invest. New Drugs, 19, 41 (2001). 55. J. G. Kuhn, Ann. Pharmacother., 35, 217 (2001). 56. P. M. G. Hoff, Semin. Oncol., 30, 88 (2003). 57. T. Watanabe, Breast Cancer, 20, 302 (2013). 58. S. Hillers, R. A. Zhuk, M. Lidaks and A. Zidermane, GB 1168391, 1969; Chem. Abstr., 72, 1170 43715 (1970). 59. S. Hillers, M. Lidaks, R. A. Zhuk, A. Berzina, K. Pets, I. N. Getsova and E. I. Bruk, Chem. Heterocycl. Compds., 5, 283 (1969). 60. C. G. Kruse, F. L. Jonkers, V. Dert and A. Van der Gen, Recl. Trav. Chim. Pays-Bas, 98, 371 (1979). 1175 61. S. Ozaki, Y. Watanabe, T. Hoshiko, T. Nagase, T. Ogasawara, H. Furukawa, A. Uemura, K. Ishikawa, H. Mori, A. Hoshi, M. Iigo and R. Tokuzen, Chem. Pharm. Bull., 34, 150 (1986). 62. M. Yasumoto, A. Moriyama, N. Unemi, S. Hashimoto and T. Suzue, J. Med. Chem., 20, 1592 (1977). 63. S. Hillers, R. A. Zhuk, A. Berzina, L. Serina and A. Lazdins, DE 2357847, 1974; Chem. 1180 Abstr., 81, 120680 (1974). 64. J. Yamashita, M. Yasumoto and S. Hashimoto, Chem. Pharm. Bull., 31, 3872 (1983). 65. C. H. Lee, J. Y. Kim, W. J. Kim and Y. H. Kim, Heterocycles, 31, 211 (1990). 66. Y. Hoshide, Y. Hashimoto, H. Ogawa and S. Tasaka, FR 2355841, 1977; Chem. Abstr., 88, 121667 (1978). 1185 67. A. Kojima, Y. Jinno, S. Ike, T. Yokoyama and M. Odate, US 4256885, 1977; Chem. Abstr., 88, 136653 (1978). 68. K. Sakurai, S. Aoyagi, H. Toyofuku, M. Ohki, T. Yoshizawa and T. Kuroda, Chem. Pharm. Bull., 26, 3565 (1978). 69. I. Minami, Y. Yoshioka and H. Nomura, DE 2657709, 1977; Chem. Abstr., 87, 135396 (1977). 1190 70. K. Ishibashi, S. Ishiguro and R. Komaki, US 4159378, 1979; Chem. Abstr., 91, 20334 (1979). 71. T. Nakamura, Y. Hashimoto, Y. Hoshide, K. Suzuki and Y. Yoshida, US 4121037, 1978; Chem. Abstr., 88, 152656 (1978). 72. U. Michael, Z. Sasson and E. Schoenberger, US 4174446, 1979; Chem. Abstr., 89, 24357 (1978). 1195 73. H. Nomura, Y. Yoshioka and I. Minami, Chem. Pharm. Bull., 27, 899 (1979). 74. P. D. Schuman and R. Anderson, US 4024143, 1977; Chem. Abstr., 87, 135401 (1977). 75. T. Kametani, K. Kigasawa, M. Hiiragi, K. Wakisaka, O. Kusama, H. Sugi and K. Kawasaki, Heterocycles, 6, 529 (1977). 76. J. Kiss, US 4169201, 1979; Chem. Abstr., 88, 105398 (1978). 1200 77. T. Iwasaki, T. Nishitani, H. Horikawa and I. Inoue, Tetrahedron Lett., 22, 1029 (1981). 78. M. Matsui and T. Ogawa, US 4324895, 1982; Chem. Abstr., 87, 102600 (1977). 79. R. A. Earl and L. B. Townsend, J. Heterocycl. Chem., 9, 1141 (1972). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 79

80. W. Y. Lin, D. W. Peng, B. Wang, L. Long, C. Guo and J. Yuan, Eur. J. Org. Chem., 793 (2008). 1205 81. S. Mocellin, P. Pilati, M. Lise and D. Nitti, J. Clin. Oncol., 25, 5649 (2007). 82. M. Hoffer, R. Duschinsky, J. J. Fox and N. Yung, J. Am. Chem. Soc., 81, 4112 (1959). 83. M. Hoffer, US 2949451, 1960; Chem. Abstr., 55, 2812 (1961). 84. GB 878060, 1961; Chem. Abstr., 57, 56538 (1962). 85. A. Holy and D. Cech, Collect. Czech. Chem. Commun., 39, 3157 (1974). 1210 86. J. Brokes, H. Hrebabecky and J. Beranek, Collect. Czech. Chem. Commun., 44, 439 (1979). 87. R. Marumoto and M. Honjo, Chem. Pharm. Bull., 22, 128 (1974). 88. S. Ozaki, T. Katakami and M. Saneyoshi, Bull. Chem. Soc. Jpn., 50, 2197 (1977). 89. B. Schwarz, D. Cech, A. Holy and J. Skoda, Collect. Czech. Chem. Commun., 45, 3217 (1980). 90. H. Aoyama, Bull. Chem. Soc. Jpn., 60, 2073 (1987). 1215 91. A. Uemura, K. Nozaki, J. Yamashita and M. Yasumoto, Tetrahedron Lett., 30, 3819 (1989). 92. M. Taverna-Porro, L. A. Bouvier, C. A. Pereira, J. M. Montserrat and A. M. Iribarren, Tetra- hedron Lett., 49, 2642 (2008). 93. I. Serra, T. Bavaro, D. A. Cecchini, S. Daly, A. M. Albertini, M. Terreni and D. Ubiali, J. Mol. Catal. B-Enzym., 95, 16 (2013). 1220 94. A. Hatano, M. Kurosu, S. Yonaha, M. Okada and S. Uehara, Org. Biomol. Chem., 11, 6900 (2013). 95. I. Votruba, A. Holy, H. Dvorakova, J. Gunter, D. Hockova, H. Hrebabecky, T. Cihlar and M. Masojidkova, Collect. Czech. Chem. Commun., 59, 2303 (1994). 96. C. N. Britos, V. A. Cappa, C. W. Rivero, J. E. Sambeth, M. E. Lozano and J. A. Trelles, J. 1225 Mol. Catal. B-Enzym., 79, 49 (2012). 97. V. A. Cappa, C. W. Rivero, C. N. Britos, L. M. Martinez, M. E. Lozano and J. A. Trelles, Pro- cess. Biochem., 49, 1169 (2014). 98. T. Kawaguchi, Y. Suzuki, Y. Nakahara, N. Nambu and T. Nagai, Chem. Pharm. Bull., 33, 301 (1985). 1230 99. Z. Zhang, H. Hatta, K. Tanabe and S. Nishimoto, Pharm. Res., 22, 381 (2005). 100. Y. Ge, X. Wu, D. Zhang and L. Hu, Bioorg. Med. Chem. Lett., 19, 941 (2009). 101. K. Yoshida, US 0111852 2010; Chem. Abstr., 149, 77397 (2008). 102. Y. Kinoshita, M. Sugimoto and T. Ishiguro, EP 2275135, 2009; Chem. Abstr., 151, 423612 (2009). 1235 103. Y. M. Wei, Y. F. Ya, D. H. Pei and B. Gong, Bioorg. Med. Chem. Lett., 8, 2419 (1998). 104. J. V. Voorde, S. Liekens, C. McGuigan, P. G. S. Murziani, M. Slusarczyk and J. Balzarini, Biochem. Pharmacol., 82, 441 (2011). 105. A. Usier-Brosceta and J. Weiss, PCT Int. Appl. WO 202994, 2014; Chem. Abstr., 162, 122992 (2014). 1240 106. R. D’Sousa and J. Kiss, US 4340729, 1982; Chem. Abstr., 95, 62602 (1981). 80 Ferraboschi, Ciceri, and Grisenti

107. J. Kiss, R. D’Souza, J. A. Van Koeveringe and W. Arnold, Helv. Chim. Acta, 65, 1522 (1982). 108. G. Bertolini and M. Frigerio, PCT Int. Appl. WO 040184, 2005; Chem. Abstr., 142, 411592 (2005). 109. H. Hrebabecky and J. Beranek, Nucleic Acids Res., 5, 1029 (1978). 1245 110. A. F. Cook, M. J. Holman, M. J. Kramer and P. W. Trown, J. Med. Chem., 22, 1330 (1979). 111. G. Cotticelli, G. De Meglio, S. Monciardini and G. Ordanini, US 0128467, 2002; Chem. Abstr., 130, 38646 (1998). 112. T. Bavaro, S. Rocchietti, D. Ubiali, M. Filice, M. Terreni and M. Pregnolato, Eur. J. Org. Chem., 1967 (2009). 1250 113. A. Hoshi and J. Castaner, Drugs Future, 21, 358 (1996). 114. G. Pentheroudakis and C. Twelves, Clin. Colorectal Cancer, 2, 16 (2002). 115. J. J. Hwang and J. L. Marshall, Expert Opin. Pharmacother., 3, 733 (2002). 116. C. M. Walko and C. Lindley, Clin. Ther., 27, 23 (2005). 117. C. Serdjebi, G. Milano and J. Ciccolini, Expert Opin. Drug Metab. Toxicol., 11, 665 (2015). 1255 118. A. Bronckaers, F. Gago, J. Balzarini and S. Liekens, Med. Res. Rev., 29, 903 (2009). 119. H. W. M. van Laarhoven and C. J. A. Punt, Eur. J. Gastroen. Hepat., 16, 283 (2004). 120. J. K. McGavin and K. L. Goa, Drugs, 61, 2309 (2001). 121. B. R. Hirsch and S. Y. Zafar, Cancer. Manag. Res., 3, 79 (2011). 122. J. Quidde, D. Arnold and A. Stein, Clin. Med. Insights Oncol., 6, 363 (2012). 1260 123. F. Y. Lee, R. Borzilleri, C. R. Fairchild, A. Kamath, R. Smykla, R. Kramer and G. Vite, Can- cer Chemother. Pharmacol., 63, 157 (2008). 124. S. Bhushan and C. M. Walko, Ann. Pharmacother., 42, 1252 (2008). 125. P. Pronzato, Drugs, 68, 139 (2008). 126. J. Bonneterre and N. Penel, Expert Opin. Pharmacother., 9, 2901 (2008). 1265 127. C. G. Murphy and A. D. Seidman, Clin. Breast Cancer, 9, S58 (2009). 128. R. J. Herr, “Capecitabine (Xeloda): an oral chemotherapy agent” in Modern Drug Synthesis,p. 57, J. J. Li and D. S. Johnson (Eds.), John Wiley & Sons, Inc, Hoboken, New Jersey, 2010. 129. H. M. Kissman and B. R. Baker, J. Am. Chem. Soc., 79, 5534 (1957). 130. G. Wang, R. C. Tam, E. Gunic, J. Du, J. Bard and B. Pai, J. Med. Chem., 43, 2566 (2000). 1270 131. J. Lee, G. S. Park, W. K. Yang, J. H. Kim, C. H. Park, Y.-H. An, C.-J. Choi, Y.-K. Chang and G. S. Lee, PCT Int. Appl. WO 105593, 2008; Chem. Abstr., 149, 308065 (2008). 132. P. Sairam, R. Puranik, B. S. Rao, P. V. Swamy and S. Chandra, Carbohydr. Res., 338, 303 (2003). 133. S. Mohanty, R. Bera, M. B. Murlidhar, T. N. Rao, K. K. Nerurkar, D. Mitra and B. P. R. Rao, 1275 IN 2010/CH/02474, 2013; Chem. Abstr., 158, 475020 (2013). 134. N. Shimma, I. Umeda, M. Arasaki, C. Murasaki, K. Masubuchi, Y. Kohchi, M. Miwa, M. Ura, N. Sawada, H. Tahara, I. Kuruma, I. Horii and H. Ishitsuka, Bioorg. Med. Chem., 8, 1697 (2000). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 81

135. C. R. Roberts and J.-W. Wong, PCT Int. Appl. WO 063786, 2005; Chem. Abstr., 143, 60196 1280 (2005). 136. G. J. B. Ettema, US 0300399, 2008; Chem. Abstr., 149, 576806 (2008). 137. R. V. Palle, A. M. Marathe, S. Aluru, R. Bochha, R. Kadaboina, S. M. Nariyam and A. Patni, PCT Int. Appl. WO 131062, 2008; Chem. Abstr., 149, 471763 (2008). 138. J. Lee, G.-S. Park, W. K. Yang, J. H. Kim, C. H. Park, Y.-H. An, Y. J. Lee, Y.-K. Chang and 1285 G. S. Lee, PCT Int. Appl. WO 066892, 2009; Chem. Abstr., 151, 8871 (2009). 139. J. Bautista Rodriguez, E. Ravaschino and E. Elhalem, PCT Int. Appl. WO 071726, 2009; Chem. Abstr., 151, 57090 (2009). 140. M. N. R. Arasaki, H. Ishitsuka, I. Kuruma, M. Miwa, C. Murasaki, N. Shimma and I. I. H. Umeda, US 5472949, 1995; Chem. Abstr., 122, 240352 (1995). 1290 141. H. R. Brinkman, P. Kalaritis and J. F. Morrissey, US 5476932, 1995; Chem. Abstr., 124, 233026 (1996). 142. P. L. MacDonald, P. Rossetto and M. Gallina, PCT Int. Appl. WO 088989, 2009; Chem. Abstr., 151, 173777 (2009). 143. K.-C. Lin and C. Chien, US 0184451, 2013; Chem. Abstr., 159, 197129 (2013). 1295 144. V. G. Gore, L. N. Patkar, R. Bhalerao, M. G. Hublikar and K. S. Pokharkar, PCT Int. Appl. WO 104540, 2011; Chem. Abstr., 155, 352803 (2011). 145. V. G. Gore, L. Patkar, A. Bagul and M. Edake, PCT Int. Appl. WO 067588, 2011; Chem. Abstr., 155, 12285 (2011). 146. B. Shen and T. F. Jamison, Org. Lett., 14, 3348 (2012). 1300 147. M. R. Dhananjeyan, J. Liu, C. Bykowski, J. A. Trendel, J. G. Sarver, H. Ando and P. W. Erhardt, J. Chromatogr. A, 1138, 101 (2007). 148. R. Kadaboina, S. M. Nariyam, V. Murki, R. R. Vinjamuri, S. Benda, S. K. Komati and N. Gunda, PCT Int. Appl. WO 065586, 2010; Chem. Abstr., 153, 11944 (2010). 149. T.-C. Hu and H.-T. Huang, PCT Int. Appl. WO 010967, 2011; Chem. Abstr., 154, 158730 1305 (2011). 150. S. Rocchietti, M. Terreni and M. Pregnolato, PCT Int. Appl. WO 057894, 2003; Chem. Abstr., 139, 116337 (2003). 151. M. Pregnolato, M. Terreni, D. Ubiali and T. Bavaro, PCT Int. Appl. WO 107771, 2008; Chem. Abstr., 149, 332574 (2008). 1310 152. T. Bavaro, D. Ubiali, S. Brocca, S. Rocchietti, I. Nieto, M. Pregnolato, M. Lotti and M. Ter- reni, Biocatal. Biotransform., 28, 108 (2010). 153. S. Ciceri, P. Ciuffreda, P. Grisenti and P. Ferraboschi, Tetrahedron Lett., 56, 5909 (2015). 154. M. Pascual Gilabert, V. M. Deroncele Thomas and R. Montilla Arevalo, EP 2883959, 2015; Chem. Abstr., 163, 103800 (2015). 1315 155. X. Jia, X. Liu, J. Wang, M. Wang, H. Guo and M. Liu, Chem. Res. Chin. Univ., 31, 78 (2015). 156. P. D. Davis, M. A. Naylor, P. Thomson, S. A. Everett, M. R. L. Stratford and P. Wardman, PCT Int. Appl. WO 032921, 2006; Chem. Abstr., 144, 350926 (2006). 157. V. Jhansi Rani, A. Raghavendra, P. Kishore, Y. Nanda Kumar, K. Hema Kumar and K. Jaga- deeswarareddy, Eur. J. Med. Chem., 54, 690 (2012). 82 Ferraboschi, Ciceri, and Grisenti

1320 158. S. Nakai, K. Sasai and T. Ezawa, Chem. Pharm. Bull., 40, 2568 (1992). 159. T. Kamiya, M. Ishiduka and H. Nakajima, US 5453497, 1995; Chem. Abstr., 122, 188034 (1995). 160. M. Kurono, T. Chiba and S. Fujii, US 4124765, 1978; Chem. Abstr., 84, 17405 (1976). 161. J. Yamashita, I. Yamawaki, S. Ueda, M. Yasumoto, N. Unemi and S. Hashimoto, Chem. 1325 Pharm. Bull., 30, 4258 (1982). 162. S. Ozaki, Y. Watanabe, T. Hoshiko, T. Nagase, T. Ogasawara, H. Furukawa, A. Uemura, K. Ishikawa, H. Mori, A. Hoshi, M. Iigo and R. Tokuzen, Chem. Pharm. Bull., 34, 150 (1986). 163. S. Fujii, US 4864021, 1989; Chem. Abstr., 105, 97331 (1986). 164. M. Hirohashi, M. Kido, Y. Yamamoto, Y. Kojima, K. Jitsukawa and S. Fujii, Chem. Pharm. 1330 Bull., 41, 1498 (1993). 165. R. A. Persichetti and J. F. Liu, PCT Int. Appl. WO 109274, 2011; Chem. Abstr., 155, 423410 (2011). 166. S. Ozaki, Y. Ike, H. Mizuno, K. Ishikawa and H. Mori, Bull. Chem. Soc. Jpn., 50, 2406 (1977). 1335 167. C. Heidelberger, D. G. Parsons and D. C. Remy, J. Med. Chem., 7, 1 (1964). 168. A. Zaniboni, P. Bertocchi, S. Barni and F. Petrelli, Clin. Colorectal Cancer (2016) DOI:10.1016/j.clcc.2016.06.003 169. R. Daifuku and D. S. Sergueev, PCT Int. Appl. WO 057828, 2016; Chem. Abstr., 164, 492751 (2016). 1340 170. K. W. Pankiewicz, Carbohydr. Res., 327, 87 (2000). 171. W.-D. Meng and F.-L. Qing, Curr. Top. Med. Chem., 6, 1499 (2006). 172. P. Liu, A. Sharon and C. K. Chu, J. Fluor. Chem., 129, 743 (2008). 173. X. L. Qiu, X. H. Xu and F. L. Qing, Tetrahedron, 66, 789 (2010). 174. V. E. Marquez, C. K. H. Tseng, H. Mitsuya, S. Aoki, J. A. Kelley, H. Ford, J. S. Roth, S. 1345 Broder, D. G. Johns and J. S. Driscoll, J. Med. Chem., 33, 978 (1990). 175. G. M. Blackburn, D. A. England and F. Kolkmann, J. Chem. Soc., Chem. Commun., 930 (1981). 176. J.-L. Paparin, E. Badaroux, C. Pierra and C. B. Dousson, WO 081297, 2015; Chem. Abstr., 163, 66114 (2015). 1350 177. K. Fauster, M. Hartl, T. Santner, M. Aigner, C. Kreutz, K. Bister, E. Ennifar and R. Micura, ACS Chem. Biol., 7, 581 (2012). 178. J. Salon, J. Sheng, J. Gan and Z. Huang, J. Org. Chem., 75, 637 (2010). 179. J. L. Clark, L. Hollecker, J. C. Mason, L. J. Stuyver, P. M. Tharnish, S. Lostia, T. R. McBrayer, R. F. Schinazi, K. A. Watanabe, M. J. Otto, P. A. Furman, W. J. Stec, S. E. Patter- 1355 son and K. W. Pankiewicz, J. Med. Chem., 48, 5504 (2005). 180. A. B. Eldrup, M. Prhavc, J. Brooks, B. Bhat, T. P. Prakash, Q. Song, S. Bera, N. Bhat, P. Dande, P. D. Cook, C. F. Bennett, S. S. Carroll, R. G. Ball, M. Bosserman, C. Burlein, L. F. Colwell, J. F. Fay, O. A. Flores, K. Getty, R. L. LaFemina, J. Leone, M. MacCoss, D. R. McMasters, J. E. Tomassini, D. Von Langen, B. Wolanski and D. B. Olsen, J. Med. Chem., 1360 47, 5284 (2004). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 83

181. K. Brown, M. Dixey, A. Weymouth-Wilson and B. Linclau, Carbohydr. Res., 387, 59 (2014). 182. L. W. Hertel, J. S. Kroin, J. W. Misner and J. M. Tustin, J. Org. Chem., 53, 2406 (1988). 183. T. S. Chou, P. C. Heath, L. E. Patterson, L. M. Poteet, R. E. Lakin and A. H. Hunt, Synthesis, 565 (1992). 1365 184. R. B. Polturi, S. H. Venkata and R. Betini, PCT Int. Appl. WO 095430, 2005; Chem. Abstr., 143, 367533 (2005). 185. R. B. Potluri, H. Venkata Subramanian, R. Betini and S. Gunturu, PCT Int. Appl. WO 095359, 2006; Chem. Abstr., 145, 315226 (2006). 186. B. Parthasaradhi Reddy, K. Rathnakar Reddy, R. Raji Reddy, D. Muralidhara Reddy and I. 1370 Srinivas Reddy, PCT Int. Appl. WO 015257, 2007; Chem. Abstr., 146, 229557 (2007). 187. M.-S. Kim, Y.-J. Kim, J.-H. Choi, H.-G. Lim and D.-W. Cha, PCT Int. Appl. WO 069838, 2007; Chem. Abstr., 147, 95870 (2007). 188. Y. Xu, H. Yang and W. Hou, US 0179314, 2010; Chem. Abstr., 148, 168918 (2008). 189. V. R. A. Palle, S. M. Nariyam, V. Murki, A. A. Waghmare and V. N. Mundhada, PCT Int. 1375 Appl. WO 117760, 2007; Chem. Abstr., 147, 449039 (2007). 190. G. C. Maikap, D. Bhatt and B. K. Panda, PCT Int. Appl. WO 092808, 2006; Chem. Abstr., 145, 293299 (2006). 191. A. Ujagare, D. A. Kochrekar and M. C. Uzagare, PCT Int. Appl. WO 049294, 2007; Chem. Abstr., 146, 468552 (2007). 1380 192. S.-J. Park, C.-R. Oh and Y.-D. Kim, PCT Int. Appl. WO 117955, 2008; Chem. Abstr., 149, 378977 (2008). 193. B. Vishnukant, P. Purohit, K. Paparao and V. Veereshappa, PCT Int. Appl. WO 026222, 2008; Chem. Abstr., 148, 331958 (2008). 194. X. R. Jiang, J. F. Li, R. X. Zhang, Y. Zhu and J. S. Shen, Org. Process Res. Dev., 12, 888 1385 (2008). 195. L. Zelikovitch, O. Friedman, T. Fizitzky and J. Manascu, US 0262215, 2008; Chem. Abstr., 149, 471761 (2008). 196. A. Varalli, P. Anesa, M. Argese and G. Guazzi, US 0105887, 2010; Chem. Abstr., 152, 493457 (2010). 1390 197. C.-Y. Chu, W.-D. Lee, W. Li and C. K. Hwang, US 0124797, 2009; Chem. Abstr., 150, 473105 (2009). 198. C. Chien, P.-S. Chien and C.-K. Hwang, US 0259106, 2012; Chem. Abstr., 157, 520692 (2012). 199. Y. K. Chang, J. Lee, G. S. Park, M. Lee, C. H. Park, H. K. Kim, G. Lee, B. Y. Lee, J. Y. Baek 1395 and K. S. Kim, Tetrahedron, 66, 5687 (2010). 200. J. Lee, G. S. Park, M. Lee, H.-J. Bang, J. C. Lee, C. K. Kim, C.-J. Choi, H. K. Kim, H. C. Lee, Y.-K. Chang and G. S. Lee, US 0249818, 2007; Chem. Abstr., 145, 124819 (2006). 201. C. Gong, US 0003963, 2006; Chem. Abstr., 144, 51839 (2006). 202. D. P. Kjell, US 5633367, 1997; Chem. Abstr., 125, 329284 (1996). 1400 203. Y. Cen and A. A. Sauve, Nucleosides Nucleotides Nucleic Acids, 29, 113 (2010). 84 Ferraboschi, Ciceri, and Grisenti

204. M. Peifer, R. Berger, V. W. Shurtleff, J. C. Conrad and D. W. MacMillan, J. Am. Chem. Soc., 136, 5900 (2014). 205. H. Zhong, A. Lv, R. Kalidindi, S. Lin and X. Su, PCT Int. Appl. WO 112473, 2007; Chem. Abstr., 147, 427652 (2007). 1405 206. J. Pulido, A. J. Sobczak, J. Balzarini and S. F. Wnuk, J. Med. Chem., 57, 191 (2014). 207. D. M. Bender and D. M. Remick, PCT Int. Appl. WO 065525, 2006; Chem. Abstr., 145, 76625 (2006). 208. S. C. Srivastava, S. K. Bajpai and K.-H. Sit, US 0159375, 2005; Chem. Abstr., 143, 133642 (2005). 1410 209. R. G. L. Shorr, R. J. Rodriguez and L. Boteju, PCT Int. Appl. WO 143593, 2011; Chem. Abstr., 155, 649030 (2011). 210. S. Maiti, N. Park, J. H. Han, H. M. Jeon, J. H. Lee, S. Bhuniya, C. Kang and J. S. Kim, J. Am. Chem. Soc., 135, 4567 (2013). 211. A. J. Martinez, A. Pendri, R. B. Greenwald and Y. H. Choe, US 6395266, 2002; Chem. Abstr., 1415 136, 401913 (2002). 212. J. M. Harris, K. Yoon, M. D. Bentley, Z. Fang and T. Viegas, US 0273074, 2015; Chem. Abstr., 151, 174043 (2009). 213. R. Pignatello, L. Vicari, V. Pistara, T. Musumeci, M. Gulisano and G. Puglisi, Drug. Develop. Res., 71, 294 (2010). 1420 214. B. Botta, C. Fraschetti, I. D’Acquarica, F. Sacco, J. Mattay, M. C. Letzel and M. Speranza, Org. Biomol. Chem., 9, 1717 (2011). 215. I. D’Acquarica, A. Calcaterra, F. Sacco, F. Balzano, F. Aiello, A. Taﬁ, N. Pesci, G. Uccello- Barretta and B. Botta, Chirality, 25, 840 (2013). 216. J. T. Weiss, J. C. Dawson, C. Fraser, W. Rybski, C. Torres-Sanchez, M. Bradley, E. E. Patton, 1425 N. O. Carragher and A. Unciti-Broceta, J. Med. Chem., 57, 5395 (2014). 217. A. Istrate, M. Medvecky and C. J. Leumann, Org. Lett., 17, 1950 (2015). 218. J. L. Eiseman, C. Brown-Proctor, P. E. Kinahan, J. M. Collins, L. W. Anderson, E. Joseph, D. R. Hamburger, S. S. Pan, C. A. Mathis, M. J. Egorin and R. W. Klecker, Clin. Cancer Res., 10, 6669 (2004). 1430 219. K. Seki, K. Nishijima, K. Sanoki, Y. Kuge, M. Takahashi, H. Akizawa, N. Tamaki, L. I. Wiebe and K. Ohkura, Heterocycles, 77, 1307 (2009). 220. K. I. Nishijima, Y. Kuge, K. Seki, K. Ohkura, N. Motoki, K. Nagatsu, A. Tanaka, E. Tsukamoto and N. Tamaki, Nucl. Med. Biol., 29, 345 (2002). 221. A. F. Shields, J. Nucl. Med., 44, 1432 (2003). 1435 222. X. Fei, J.-Q. Wang, K. D. Miller, G. W. Sledge, G. D. Hutchins and Q.-H. Zheng, Nucl. Med. Biol., 31, 1033 (2004). 223. B. S. Moon, A. Y. Shim, K. C. Lee, H. J. Lee, B. S. Lee, G. I. An, S. D. Yang, D. Y. Chi, C. W. Choi, S. M. Lim and K. S. Chun, Bull. Korean Chem. Soc., 26, 1865 (2005). 224. D. O’Hagan and C. Schaffrath, PCT Int. Appl. WO 020945, 2003; Chem. Abstr., 138, 237025 1440 (2003). Synthesis of Antitumor Fluorinated Pyrimidine Nucleosides 85

225. L. Martarello, C. Schaffrath, H. Deng, A. D. Gee, A. Lockhart and D. O’Hagan, J. Label. Compd. Radiopharm., 46, 1181 (2003). 226. D. O’Hagan, J. Naismith, C. Schaffrath, C. Dong, J. B. Spencer and F. Huang, PCT Int. Appl. WO 078914, 2004; 141, 256535 (2004). 1445 227. M. Winkler, J. Domarkas, L. F. Schweiger and D. O’Hagan, Angew. Chem. Int. Ed., 47, 10141 (2008). 228. M. Onega, J. Domarkas, H. Deng, L. F. Schweiger, T. A. Smith, A. E. Welch, C. Plisson, A. D. Gee and D. O’Hagan, Chem. Commun., 46, 139 (2010). 229. D. O’Hagan and H. Deng, Chem. Rev., 115, 634 (2015). 1450 230. L. Ma, Y. F. Li, L. P. Meng, H. Deng, Y. Y. Li, Q. Zhang and A. P. Diao, RSC Adv., 6, 27047 (2016). 231. H. Wojtowicz-Rajchel, J. Fluorine Chem., 143, 11 (2012).